Abstract

Predictive-based power control has been widely recognized as a promising approach to boost driving range and improve system-level energy efficiency for electric vehicles (EVs), in which vehicle velocity forecasting generally serves as a preliminary input to optimally schedule the operations of varying onboard electrical and thermal systems. A segment-based velocity forecasting approach for individual commuting vehicles developed in this study reveals that it is challenging to forecast the velocity at intersection segments only using the velocity data. To address this challenge, this study seeks to develop a YOLO-V2-based object detection deep network to recognize the traffic lights in advance and leverage the detected signals to establish a forecasting model that integrates with the probability-based hybrid forecasting approach. The case study results show that the traffic light detection-based forecasting model can significantly improve the forecasting accuracy for intersection segments. Based on the forecasting velocity 5–15 s ahead, the effectiveness of model predictive control-based energy management strategy is further evaluated with a liquid-based battery thermal control system. The proposed battery thermal management system (BTMS) model shows promising results in maintaining battery temperature within an appropriate range, thus improving the overall energy efficiency of the EV. Moreover, a traffic light-based real-time energy management framework is developed to directly control the power demand from the air conditioning (AC) system.

1 Introduction

As one of the most promising approaches to address the climate change caused by massive transportation, lithium-ion battery-based vehicle electrification has emerged, developed, and advanced with an overwhelming speed during the past decade. However, there still exist several technical barriers preventing its further expansion for electric vehicles (EVs) such as battery safety, battery lifespan, and driving range anxiety. A safer battery pack with higher energy density, larger capacity, and higher reliability has always been a consistent aspiration for the fast-growing EV industry. Theoretically, these concerns can be potentially fully mitigated as battery technologies advance. At the current stage, multiple feasible solutions have been developed from cell to pack simultaneously, such as seeking high energy density electrode materials and solid state electrolytes, employing integrated cell geometry designs like cylindrical type-4680 and prismatic blade battery, and utilizing space-efficient pack assemble methods like cell-to-pack and cell-to-chassis [1]. Additionally, based on these readily available systems, it is achievable to alleviate the range concerns by improving system-level energy efficiency via predictive thermal control and optimal power allocation.

1.1 Battery and Vehicle Thermal System.

Extensive studies have shown that the performance of lithium-ion batteries is highly influenced by temperature variations. High temperatures can lead to faster degradation and potential thermal runaway, while exposure to temperatures below 0C can cause irreversible damage due to lithium plating, resulting in a significant reduction in power and capacity. State-of-the-art BTMS use a variety of heat transfer mediums to maintain appropriate temperatures, including air, liquid, phase changing materials (PCMs), and heat pipes. Early BTMS focused on developing methods to remove the waste heat from a single battery system, with active air cooling commonly used in early hybrid EV models such as the Toyota Prius and Nissan Leaf. However, after Tesla successfully integrated its liquid-based battery thermal system with the vehicle air conditioning (AC) system, research on BTMS broke into two different scopes and developed in parallel.

Regarding component-level research, the relationships between the cell and thermal medium have been mainly investigated in three topics. First, novel thermal materials like PCM, heat pipe [2], and hybrids among them have been evaluated with improved performance, but the applications of these materials are still limited to laboratory level due to their intrinsic complexity. Second, ventilation arrangements like Z, U, and J-type [3], spoiler [4], and their further advancements [5,6] have been designed and proven to have better thermal performances in terms of maximum temperature control and uniformity. Similarly, three major liquid-based structures have been developed based on different battery geometries. Third, multiple linear and nonlinear algorithms have been developed to constrain battery temperatures for daily driving. Additionally, model predictive control (MPC) that foresees several steps ahead and yields an optimized control solution has also been employed for real-time component-level dynamic controls.

At the system level, robust battery thermal control requires strong heat dissipation capacity to remove waste heat under heavy working conditions, which typically employs liquid as the heat transfer medium and is coupled with an onboard AC system or radiator. Optimizing battery and cabin thermal controls simultaneously at the system level has become a trend in EV industrial applications in recent years. Multiple control strategies have been successfully developed for liquid cooling configurations, but the complexity of multisystem modeling remains a major concern in implementing MPC for thermal control. The reported literature tends to develop nonlinear MPC models or establish hierarchical structures. For example, Amini et al. [7], Tang et al. [8], Park and Ahn [9], and Zhao and Mi [10] have proposed various MPC strategies to regulate battery and cabin thermal control. These strategies have achieved higher efficiency and better regulation of thermal requirements based on driving cycle information. The discussed predictive strategy also has the potential to be integrated with the whole power users to achieve better energy performance.

Besides thermal dissipation, thermal preservation in extremely low temperatures has garnered interest. Traditional methods like resistance heaters and heat pumps are used to maintain temperatures, but there is an increasing interest in using PCMs as thermal reservoirs, in which the heat storage enables passive thermal preservation for short-time parking and even overnight parking with higher ambient temperatures [11,12].

1.2 Vehicle Energy Management.

Minimizing energy consumption through joint optimization of vehicle energy systems, including thermal systems, is essential for EVs. Besides driving motors, the air conditioning system for both the cabin and battery is the second highest power-consuming system for EVs, followed by entertainment and other auxiliary systems such as liquid pumps for BTMS. Dynamic load shifting is a potential solution to prevent high power demand overlap, which has two benefits: (i) lower power output density mitigates cell swelling and prolongs battery cycle life [1315] and (ii) high power density generates more waste heat, which negatively affects thermal control and energy savings. Therefore, optimizing regenerative charging and discharging sequences in conjunction with thermal control systems is a reasonable approach.

Similar to the thermal control for multiple cooperated systems, besides the rule-based deterministic methods and offline optimization algorithms like Pontryagin’s minimum principle and dynamic programming, MPC is an effective online optimal control method for coordinating load shifting with battery and cabin thermal control. For example, hierarchical MPC has been used for eco-cooling of connected and autonomous vehicles [16], optimizing battery operations and discharging schedules [17], and for vehicle-level and electric powertrain-level optimizations simultaneously [18]. Learning-based algorithms, particularly reinforcement learning-based methods, have also been successfully implemented in plug-in hybrid EVs [19]. However, studies focusing on pure EVs are rare due to the complexity of nonlinear multisystem optimization [20,21]. In addition, applying these predictive algorithms in vehicles requires robust onboard computational resources for real-time processing.

1.3 Vehicle Velocity Forecasting.

In order to achieve optimal control in energy management algorithms, accurate velocity forecasting is essential. In particular, short-term local intersection prediction requires high accuracy to avoid potential instability among multiple systems. Individual vehicle velocity prediction has become increasingly important with the advent of big data and machine learning-based forecasting technologies [17,22].

Compared with network traffic forecasting that emphasizes more on traffic information in networks and aims to provide further insights for transportation management and policy-making, there are three major discrepancies between individual and network vehicle velocity prediction in general. (i) Individual vehicle velocity prediction uses floating speed trajectories as a data source rather than network traffic records. (ii) Velocity prediction for individual vehicles necessitates prediction horizons at seconds, which is much shorter than timescales in minutes or hours for network traffic prediction. (iii) For computing platforms, individual velocity forecasting has no alternative but to directly implement onboard computing tools due to real-time requirements, while network traffic forecasting can utilize local/cloud-based platforms with higher computing capability.

As discussed earlier, there are two broad categories of algorithms for individual traffic forecasting: stochastic and deterministic, with deterministic approaches being more suitable due to computing limitations. The hidden Markov model (HMM) chain algorithm is a widely used stochastic method for annotating sequential data with underlying hidden structures. Several modifications have been proposed to adopt HMM for vehicle velocity forecasting, such as using a fuzzy logistic model by Jing et al. [23] to estimate individual vehicle speed 8 s ahead, a self-learning multistep Markov chain model proposed by Zhou et al. [24] based on simulated data, and a second-order HMM model for dynamic forecasting model selection in segment-based vehicle velocity forecasting by Liu and Zhang [25]. However, studies have shown that deterministic approaches have higher prediction accuracy, especially with limited datasets. For example, Sun et al. [26] compared stochastic (HMM) and deterministic algorithms (support vector machine, radial basis function neural network, and back-propagation neural network) for short-term vehicle velocity forecasting, and the results showed that deterministic algorithms outperformed HMM in terms of prediction accuracy. Liu et al. [27] also validated similar results by evaluating long short-term memory network, auto-regressive integrated moving average, and HMM for 10 s ahead speed forecasting on a real urban driving dataset. By leveraging the strengths of both methods, it is practical to combine the deterministic and stochastic approaches. For example, Shen et al. [28] developed a long-term velocity forecasting method by integrating a transformer network with the Markov chain Monte Carlo algorithm, achieving enhanced performance compared to the particle filter and long short-term memory (LSTM) methods. Moreover, Lemieux and Ma [29] investigated a deep belief network with a stacked auto-encoder for highway speed forecasting.

It is worth noting that individual vehicle velocity forecasting can be potentially improved by considering surrounding network traffic situations such as image detection of front vehicles and reported traffic accidents [30,31]. However, the policy-making and construction of intelligent transportation systems are currently falling behind the trend of electrification and intellectualization. Vehicles have limited access to real-time traffic information given the cost of data communication, but approaching traffic information can be collected via existing vehicle installations such as radar and camera systems, and traffic light detection technologies could be used to further improve velocity forecasting. Unlike traditional approaches that treat vehicle velocity prediction as a standalone component, our study proposes a comprehensive framework that directly integrates velocity forecasting into the energy management systems of EVs. This integration enhances the coordination between vehicle-level and component-level (battery, air conditioning) frameworks, improving the responsiveness of EVs to dynamic driving conditions and contributing significantly to energy optimization.

1.4 Research Objective.

Aiming to enhance the vehicle energy efficiency, in this article, a hybrid two-stage localized model selection framework is developed for short-term vehicle velocity forecasting. Real-time detected traffic light signals are incorporated to boost velocity forecasting accuracy and optimize power allocation at intersections. Compared with the nonlinear MPC approach, the real-time traffic light detection-based energy management method achieves similar level of energy efficiency, but requires significantly less computational resources. Furthermore, a novel sandwich-like cell-to-chassis battery cooling prototype is developed and validated using transient computational fluid dynamics (CFD) simulations, in which we exploit the potentials of using PCM to store heat under cold environments. Overall, this study seeks to develop alternative energy management methods with lower computational cost rather than predictive algorithms to further enhance vehicle energy efficiency. Our contributions are threefold: (i) propose a two-stage localized model selection-assisted vehicle velocity forecasting framework integrated with image-based traffic light detection as an indicator for intersection velocity forecasting; (ii) develop a liquid-based plate cooling structure with PCM as the thermal buffer and heat reservoir; and (iii) explore the feasibility of leveraging traffic light detection for real-time intersection energy management.

The remainder of the article is organized as follows. In Sec. 2, a short-term velocity forecasting framework is developed with the integration of traffic light detection. The vehicle dynamic model and battery electric-thermal model are established in Sec. 3. Comparative studies between the MPC based and traffic signal light detection-assisted real-time energy management are conducted in Sec. 4. Finally, Sec. 5 summarizes the concluding remarks and the potential future work.

2 Short-Term Velocity Forecasting via Traffic Light Detection

2.1 Data Preprocessing of a Repeated Commuting Driving Cycle Dataset.

From the perspective of vehicle energy management, compared to a discrete inconsecutive driving dataset with multiple participants, a repeated driving cycles on a fixed route with the same driver has huge advantages serving as a fundamental spatial-temporal basis for algorithm development, emulation, and validation. Though several repeated driving datasets have been investigated in the literature [27,32], there are few datasets publicly available at the current stage due to data privacy and project constraints. To this end, we have successfully generated a Dallas repeated driving (DRD) cycle dataset2 with dozens of driving cycle tests performed on a fixed route in the Dallas area. Its main purpose is to simulate a typical commuting route for passenger vehicles, which consists of a 5-km expressway test and an urban local road test of 20 km, as illustrated in the dataset module in Fig. 1. According to the regional traffic retiming program [33], it is worth noting that a large portion of the traffic signals is timing controlled in a daily manner, meaning that a traffic light tends to follow the same patterns at the same time of a day.

Fig. 1
The framework of short-term vehicle velocity forecasting based on traffic light recognition and driving cycle segmentation
Fig. 1
The framework of short-term vehicle velocity forecasting based on traffic light recognition and driving cycle segmentation
Close modal

Prior to forecasting, a piece-wise approach that divides the whole cycles into segments is implemented in this study, in which the intersections and casual locations with high probability of stopping are primarily identified based on the driving data rather than directly using labeled data from public maps. By analyzing the DRD cycle dataset, it is found that nearly a quarter of those intersections should not to be treated as regular intersection due to a relatively low stopping probability, while multiple locations tend to result in indirect hidden stops under heavy traffic conditions. Once the intersections or stops are located, the routes in between are defined as independent segments. An intersection segment consists of a deceleration, a waiting, and a reacceleration process, while a normal road segment refers to a continuous moving at a steady speed. The details of the DRD cycle dataset and the preprocessing can be found in our previous study [25].

2.2 Localized Hybrid Model for Short-Term Velocity Forecasting.

Building on the segmentation approach, a forecasting framework is constructed with a two-stage structure to perform the short-term velocity forecasting. In Stage I, a forecasting submodel pool that consists of a collection of stochastic and deterministic methods are established based on their popularity, i.e., LSTM, artificial neural network (ANN), support vector regression (SVR), HMM trained with augmented data, and a similarity-based method. Their kernels, training algorithms, and hyperparameters are tuned after training and validation. This process utilizes 23 cycles (cycle 1–23) for testing and 5 cycles (cycle 24–28) for validation.

The proposed similarity-based method utilizes the similarities among historical driving cycle segments, which takes multiple previous steps and the real-time position as the inputs to retrieve the most similar historical sequences near a specific location. The similarity is calculated based on the Euclidean distance given by
(1)
where S denotes the sequence to forecast, T denotes the historical sequences used for similarity calculation, and n denotes their length. The parameter α is a set of unequal weighted factors assigned to different steps, in which the latest step has a larger weight. By adding up the most similar weighted historical sequences, we can obtain forecast speeds at different lead times. The output of stage I is an ensemble forecast produced by combining the outputs of each method in the submodel pool.

In stage II, the hybrid approach integrates the ensemble forecast from stage I with two different models: the offline probability-aided ensemble model and the online Markov chain model with dynamic model selection. This is done to mitigate potential fluctuations and uncertainties, thereby improving the accuracy of short-term velocity forecasting.

The probability-aid averaging model takes advantage of the historical statistic model ranking and directly ensembles the best three submodels for a specific segment.
(2)
where Mi denotes the ith predictive model in the ensemble. Each model, Mi, is selected based on its demonstrated performance in vehicle velocity forecasting. N represents the total number of models incorporated in the ensemble, facilitating a comprehensive approach by leveraging a variety of predictive techniques. ωi denotes an equal weighting factor with a generalization purpose for all the divided segments, s.t. ωi=1. On the other hand, the Markov chain-based model takes into account the top two models in the previous two states/segments [St2,St1,St], where St=[Mti,Mtj], and i,j[1,2,3,4,5]. The newly adopted models are determined by the maximum likelihood and are highly related to the performance of the previous two models as illustrated in Fig. 2 and expressed as follows:
Fig. 2
The dynamic model selection sketch using second-order Markov chain model
Fig. 2
The dynamic model selection sketch using second-order Markov chain model
Close modal
(3)

During the training process, eight scenarios are explored and three conditions for dynamic model selection are taken into account, followed by an ensemble step that is similar to the offline approach. A larger weight factor is assigned to the chain sequences with top models. It is also worth noting that due to the intrinsic localized characteristics, unique transition matrices are established for each segment rather than using a universal transition mechanism for the whole driving cycle. The biggest difference between these two hybrid models lies in the required model numbers for each segment, in which the universal transition mechanism needs a real-time online ranking among all the models, while the unique transition matrices can update the models offline based on the recorded velocity data. Stage II reuses the stage I validation dataset combining a collection of new data (cycle 29–31) for training, while the rest cycles (cycle 32–34) are used for testing.

Performance comparisons among the forecasting methods are presented in Table 1, where the mean absolute error (MAE) and root-mean-square error (RMSE) matrices are employed for evaluation.

Table 1

Comparisons among different forecasting methods for 5/10/15 s ahead forecasting

Lead time5 s10 s15 s
ModelMAERMSEMAERMSEMAERMSE
ANN1.221.882.073.112.503.61
LSTM1.341.972.093.002.363.37
Prob. averaged1.161.871.892.932.263.34
MM averaged1.141.851.872.912.253.44
Lead time5 s10 s15 s
ModelMAERMSEMAERMSEMAERMSE
ANN1.221.882.073.112.503.61
LSTM1.341.972.093.002.363.37
Prob. averaged1.161.871.892.932.263.34
MM averaged1.141.851.872.912.253.44

Note: Prob. averaged represents probability-aided ensemble model averaged, and MM averaged represents Markov chain model averaged. Bold values indicate the best MAE or RMSE within each category.

It can be observed from Table 1 that the proposed hybrid models show promise for improving short-term forecasting. The Markov chain model-based method with online dynamic model selection tends to yield slightly better results than the offline probability method, but the latter is more likely to be embedded into real industrial applications, especially for a shorter lead time, due to model update frequency and onboard computational resources. In the following sections, the probability-based method will be employed.

Besides the hybrid model attempts, we also recognize that the velocity profiles at intersections can be generally classified into three groups, i.e., moving forward at a constant velocity, passing through with deceleration, and completely stop, as illustrated in the intersection velocity model in Fig. 1. As a popular data mining technique, extensive surveys have implied that forecasting can be significantly enhanced via accurate classification. A traditional physical model-based classification method is developed in this study to divide the velocity sequences with unequal length, which directly utilizes the predefined deceleration, reacceleration, and stop processes as a classification threshold. We have also tested the hierarchical classification algorithm integrated with dynamic time warping and obtained similar results. Here, the prediction window is set as 5 s to align with the intersection dynamics like traffic light settings, average speed, and traffic volumes at Dallas area. Another consideration is to cooperate with control interval of potential predictive algorithms. Using the base models to forecast the velocity 5 s ahead, an averaged improvement of 0.11m/s regarding MAE is observed for all the intersection segments compared to an original MAE of 1.80m/s. Possible reasons of achieving such a limited enhancement can be attributed to the dataset size and the sample size unbalance among varying groups, which also means that the limitations can be mitigated by enlarging the dataset. Owing to the energy management focus in this study, we only present the basic flow path and summarize the takeaways that may be beneficial to future in-depth studies. Substantial explanations can be found in our previous study [25].

2.3 Short-Term Velocity Forecasting via Traffic Light Detection.

As discussed earlier, intersection velocity classification is a promising effort to improve velocity forecasting. However, when it comes to real forecasting practice, it is extremely challenging to classify the unknown future velocity sequence by merely using the velocity data. For the majority of situations, we notice that the aforementioned three scenarios at intersections are strongly associated with the traffic light signals, as presented in intersection velocity model. Moreover, it is observed that a green light usually leads to a steady running with high velocity, while the vehicle tends to stop at red/yellow lights. For the deceleration and reacceleration scenario, it can either be a green light at heavy traffic conditions or an ending red light followed by a green light. Given this consideration, in this study, we aim to develop an image-based indication framework for velocity forecasting by detecting and identifying the extra traffic control light signals via object detection. Building on this, a cornerstone of our proposed energy management strategy is the innovative use of traffic light detection. This feature significantly refines our vehicle velocity forecasts, particularly at intersections, a critical point for energy consumption in urban driving. By integrating real-time traffic light data, our system offers more precise control over the EV’s energy resources, leading to marked improvements in efficiency and a reduction in unnecessary power usage.

Extensive studies have been conducted on traffic light detection by leveraging novel and more effective convolutional neural networks (CNNs) in the emerging field of autonomous driving, covering a broad range of deep learning structures such as the R-CNN family (fast R-CNN, faster R-CNN, and mask R-CNN), the YOLO family (v1-v5), the single shot detection family, and the Retina-net family [34]. Since the motivations of this study emphasize on the impact of traffic light detection to vehicle velocity forecasting, a one-stage YOLO-v2 network with pretrained network structures is directly employed here after modification. Compared with other networks, YOLO-V2 possesses an effective classification backbone with 19 convolutional layers and 5 max-pooling layers, which provides an accurate detecting precision while maintaining a high processing rate.

For this DRD cycle dataset, all the traffic lights are horizontally installed with the same light arrangements for different colors, making it possible to directly detect the traffic light and its corresponding colors. An approximate of 1100 images are extracted from the driving cycles for labeling. To prevent potential recognition errors during the cycle, we not only label a single group of lights but also combine and label two nearby groups of lights as a whole. As an outcome, a traffic light and its color are recognized only when both objects have positive feedbacks. Our methodology employs a YOLO-V2-based object detection model, which stands out for its compatibility with standard automotive camera systems. This choice strategically avoids the high costs associated with LiDAR technologies, aiming to democratize advanced energy management capabilities in EVs. The practicality of this approach not only reduces the implementation costs but also facilitates easier adoption across various EV platforms. The model received an averaged precision of 0.845 and an averaged recall of 0.567. Compared to the typical accuracy around 0.9 reported in the literature [35], this basic model here in this study still needs further improvements.

The possible reasons and potential approaches to further improve the traffic light detection accuracy are twofold: (i) We empirically labeled all the vague images for the forecasting-oriented purposes to identify traffic lights as earlier as possible, which brings in a large amount of misleading noise. (ii) The images in this study were taken in a 3X optical zoom by a household camera device, making the images in an undesirable low quality. However, we still confidently received some inspiring results in traffic light detection: the model is able to detect the traffic light with its correct color at an approximately 100 m away in a straight road with no slopes, which is also 5–6 s ahead the intersection at a constant cruising speed. It is worth noting that the confidence score threshold in this study is set as 0.38 as a tradeoff between detection accuracy and exploration. Although there may be potentially misidentifications, i.e., detect other objects as a single group of traffic light, the object detection model is still able to yield desirable outcomes with the implementation of labeling near groups of lights.

Given the intersection velocity model and classification discussed earlier, the detected traffic light signals can act as a classification and a model indicator for the coming short-term velocity forecasting. Three different scenarios and their corresponding models are predefined according to the indicator outcomes, as demonstrated in Algorithm 1. For the sake of simplification, we only use ANN to establish these models. Especially, we find out that the reacceleration delay is highly related to the stopping location, which can be empirically calculated by a sum of the drivers’ reaction time:
(4)
where D represents the distance between the vehicle and the intersection. Parameters α and β are the coefficients depending on drivers’ patterns and vehicle performance. Here, α and β are determined as 1.18 and 2, respectively.

Traffic light-based intersection velocity model

Algorithm 1

Scenario-1: Moving forward at a constant velocity, Model: M1

Scenario-2: Passing through with deceleration, Model: M2

Scenario-3: Completely stop, Model: M3-1 for deceleration, M3-2=0 for waiting, M3-3 for reacceleration

Inputs: Detected traffic light: {green,red,yellow}

    Velocity input: [vt1,vt],a=vtvt1

Definition:a<=1.2m/s2 deceleration ; a>=1.2m/s2 acceleration (once detected, the status will be stored)

Switch  lightcolor

Case  green

   If no red signal detected previously & deceleration detected Then Scenario-2: M2 end

   If no red signal detected previously & deceleration undetected Then Scenario-1: M1 end

   If red signal detected previously Then Scenario-3: M3-3 end

Case  yellow

   If no green signal detected previously Then Scenario-3: M3-1

   If green signal detected previously & deceleration undetected Then Scenario-1: M1 end

   If green signal detected previously & deceleration detected Then Scenario-3: M3-1 end

Case  red

   If deceleration detected Then Scenario-3: M3-1 end

   Ifvt<=1m/sThen Scenario-3: M3-2 end

    (vt<=1m/s considered as a complete stop at the traffic light)

The forecasting differences between the probability-based hybrid method and traffic light detection-assisted method are compared in Fig. 3. By comparing the 5 s ahead forecasting, it is seen that the traffic light detection-assisted method results in a significant improvement from 1.56m/s to 0.78m/s regarding MAE for this specific intersection. The improvements mainly come from the deceleration process where the red light signal acts as an indicator to directly determine the ongoing stopping scenario, while there are still unavoidable forecasting delay in the reacceleration process. For other scenarios, only small enhancements are observed by adopting the traffic light detection-assisted method, leading to a limited overall improvement when it comes to whole driving cycle, i.e., an averaged 0.02m/s improvement for MAE. Note that there is no improvement or even worse outcome for the 10 s ahead forecasting using the detection-assisted approach, since most of the deceleration and reacceleration processes occur within or around 10 s. Overall, as a promising indicator, the image-based traffic light detection can be leveraged to improve the EV energy efficiency in a twofold manner: (i) traffic light detection tends to increase the forecasting accuracy for short-term velocity forecasting and (ii) it also has the potential acting as a mode trigger to activate or terminate the functions and operations of devices in advance in a predictive energy management strategy. Overall, the multiperiod short-term velocity results shown in Fig. 4 are sufficient for energy management.

Fig. 3
A comparison between the traffic light detection-based method and the hybrid forecasting method at an intersection (labels 1–3 mark the improved sections, and label 4 indicates a worsen section for 10 s ahead forecasting)
Fig. 3
A comparison between the traffic light detection-based method and the hybrid forecasting method at an intersection (labels 1–3 mark the improved sections, and label 4 indicates a worsen section for 10 s ahead forecasting)
Close modal
Fig. 4
Multiperiod velocity forecasting for the whole driving cycle
Fig. 4
Multiperiod velocity forecasting for the whole driving cycle
Close modal

3 Vehicle System Modeling

There are various systems integrating and working as a whole in an EV from the perspective of electric, thermal, and energy control, including the main battery system, vehicle motor system, air conditioning system, battery thermal control system, and the cooling functions for varying components, i.e., AD–DC bidirectional inverter, DC–DC converter, and driving motor. The section attempts to develop control-oriented models for the aforementioned systems, aiming to provide a comprehensive overview for further energy management.

3.1 Vehicle Battery System.

For an EV, the battery packs serving as the only energy source are responsible for providing sufficient energy outputs to satisfy the power demand from varying devices and subsystems. At the battery pack level, based on the first-order lumped equivalent circuit model, the effective power output from the battery pack to the power bus can be calculated as follows:
(5)
where Vocv denotes the open-circuit voltage and Rb=Ri represents the total internal resistance of a battery pack. The internal resistance R is dynamically determined, which highly depends on the operating temperature T, state of charge (SoC), and current I, and can be extracted using the hybrid pulse power characterization [36]. For the sake of simplification, within a short time of period at a steady operating condition, these two variables can be treated as a fixed constant. Equation (5) is regarded as a quadratic equation with only one variable, and the current can be calculated as follows:
(6)
From users’ perspective, the total effective power output also equals to the sum of all the power demands from subsystems, as given by:
(7)
where Pdrv, Paux, Pac, and Pbs are the power for driving motor, auxiliary devices, air conditioning system, and battery thermal control system, respectively.
Regarding the open-circuit voltage Vocv, it has a positive correlation with SoC, which can also be estimated using a polynomial regression model:
(8)
where P is a 1×7 regression matrix that relates the open-circuit voltage Vocv to the SoC through a polynomial regression model. For a specific lithium manganese oxide battery, the values of P are given by [40.2,138.6,186.2,123.5,42.4,7.5,3.3], as determined in the study conducted in Ref. [37]. The state of charge vector Ssoc is defined as Ssoc=[SoC6SoC2,SoC,1]T.
The battery SoC changes as the battery discharges and charges dynamically, as defined using the Coulomb counting by
(9)
where Qc is the original nominal battery capacity.
Moreover, from the perspective of energy balance, it is claimed that the waste energy or power loss in the form of sensible heat can be attributed to the internal resistances, expressed as follows:
(10)
where G˙ and V are the volumetric heat generation rate and total battery volume, respectively. Based on the raw experimental data, the electrothermal model is established using a surrogate-based Kriging model, and the details of the thermal model can be found in our previous study [3].

3.2 Vehicle Dynamic System.

For a running vehicle, the equivalent traction power consists of four major terms: the rolling friction, gravitational potential, air friction, and vehicle acceleration, defined as follows:
(11)
where μ=0.01 and Cd=0.24 are the rolling resistance and air friction coefficient, respectively. The windward area of a vehicle Af is set as 2.22m/s, and the vehicle mass m equals to 1875 kg for a specific passenger vehicle. The parameter h denotes the altitude, which is nonnegligible for an urban driving cycle with varying terrain and highway overpasses. It is also revealed that the motor power efficiency ηm is highly depending on motor rotating speed and torque output, ranging from 0.8 to 0.95 for varying conditions [38]. Here, ηm is also set as a constant with an average efficiency of 0.91.
When in deceleration or downhill driving conditions, the regenerative braking system is activated to harvest the extra kinetic energy with a predefined averaged efficiency back to the power bus, i.e., 80% (ηr) of the kinetic energy in this study. The driving power is negative when regeneration occurs according to the definition in Eq. (4), expressed as follows:
(12)

The regenerative energy may either be utilized directly for AC and BTMS or be recharged back to the battery pack. How and when to optimally distribute the regenerative energy needs prompt solutions, which is also one of the major contributions in this study.

3.3 Air Conditioning System.

Compared to conventional internal combustion engine (ICE) vehicles, the air conditioning system in EVs differs in multiple perspectives: (i) in winter conditions, a heat pump-based AC system for EVs has predominant advantages on energy efficiency toward conventional AC with positive temperature coefficient (PTC) heaters, while ICE can directly utilize the waste heat from the engine without any auxiliary devices or equipment; (ii) compared with ICE, there are several extra thermal loads for EVs, such as battery thermal control, motor cooling, and converter cooling; and (iii) the AC compressor in ICE is propelled by the engine, while it is powered by the battery pack for EVs. All these unique characteristics lead to a heat pump-based AC with more complicated structures and larger cooling/heating capacity for EVs, as illustrated in Fig. 5. Note that here are two types of heat sinks designed for different operating and environmental conditions separately, i.e., the AC-based chiller deals with intense driving and charging conditions under high temperatures, while the radiator usually works at low ambient temperatures. This study mainly focuses on the impacts of AC operations.

Fig. 5
A simplified AC cooling mode for EVs (1: BTMS pump, 2: BTMS heat exchanger/evaporator, 3: battery pack, 4: AC evaporator, 5: compressor, 6: BTMS three-way value, 7: condenser, 8: BTMS radiator, 9: cooling fan/blower for different systems, and 10: expansion valve)
Fig. 5
A simplified AC cooling mode for EVs (1: BTMS pump, 2: BTMS heat exchanger/evaporator, 3: battery pack, 4: AC evaporator, 5: compressor, 6: BTMS three-way value, 7: condenser, 8: BTMS radiator, 9: cooling fan/blower for different systems, and 10: expansion valve)
Close modal
A dozen of simplified conventional AC models have been developed in the literature using component-based dynamic thermal-fluid models or energy-based mathematical methods [39]. For vehicle energy evaluation, linear control-oriented models have been constructed using numerical approaches [7,40]. In our previous study [17], a linear mathematical AC model with only cooling mode was established based on the simplified AC model developed by Pino et al. [41] and the simulink-based coolsim platform developed by NREL [42]. This control-friendly model consists of three parts, i.e., an AC cooling/heating capacity model, a vehicle cabin thermal load model, and an AC energy control model. From the view of energy, the relationship between the overall power consumption and the cooling/heating capacity provided can be modeled using the coefficient of performance (CoP) (ηcop), defined as follows:
(13)
where Qac and Pac are the cooling/heating capacity and the power consumed, respectively. For a specific AC system, the CoP highly depends on the operating parameters, including the internal temperature, the external temperature, and the partial load ratio Plr. The CoP for cooling ranges from 1.8 to 4.5 when the partial load ratio is larger than 0.2, and the maximum CoP of the heating mode that can be achieved is around 1.5 at a subzero temperature, as indicated in Fig. 6. The heat pump tends to lose its advantages in energy efficiency at a 20C low temperature compared to PTC heaters [43]. It is also found that besides the two-phased refrigerant, the other heat transfer medium of the evaporator is cabin air, which is the cooling liquid for the BTMS heat exchanger. According to the refrigerant-based BTMS study conducted by Shen and Gao [44], the CoP discrepancies for different external heat transfer mediums are limited and are neglectable for energy-level modeling. In this study, the same CoP model is utilized for the sake of simplification.
Fig. 6
The coefficient of performance of the cooling mode
Fig. 6
The coefficient of performance of the cooling mode
Close modal
For the cabin thermal load model, there are several external and internal heat loads identified after considerable assumptions. The detailed modeling can be found in Ref. [17], which is also briefly described in Table 2. The transient thermal model of the vehicle cabin is formulated as follows:
(14)
where Tin(k) represents the cabin air temperature, Qac is the control variable, and δt is the time-step. The term ρairVinCair is the thermal capacity of the cabin air, while the terms Qcr, Qcw, Qr, Qh, Qf, and Qs are external and internal heat loads, as described in Table 2. Notably, one of the advantages of EV modeling at the energy level is the flexibility of the AC system in regulating thermal capacity without considering heat transfer limitations at the component level. A basic proportional integral controller can be used to maintain the cabin temperature within a comfortable zone.
Table 2

Vehicle cabin thermal modeling

Physical termSymbolheat sourceTemperatureEstimation (W)Descriptions and highlights
Conduction/convection load via roof panelQcrExternal air solar radiationDependentEq. 9 in Ref. [17]
or without radiationδ: thickness; λ: conductivity; A=3.6m2;
Qcr=kcrA(TexTin)hin=25W/(m2K); hex=4.65+13.96v
kcr=(δλ+1hin+1hex)1
Conduction/convection load via windowsQcwExternal air solar radiationDependentEq. 9 in Ref. [17]
without radiationA=1.5m2; hin=25W/(m2K);
Qcw=kcwA(TexTin)hex=4.65+13.96v
kcw=(δλ+1hin+1hex)1
Solar radiation through windowsQrSolar fluxIndependentQr=i=1n=4ηIAisinθiβFour windows: windshield, rear, left, right;
β: shading factor; θ: installation angle;
η: penetration rate; I: incident radiation
Human body thermal loadQhDriver and passenger(s)IndependentQh=145+116n [40]n: passengers number; n=3
Fresh air thermal loadQfFresh airDependentQf=ξm˙airCair(TexTin)Ventilation fresh air portion ξ=12%;
m˙air=0.186kg/s
Sensible heat loadQsCabin interiorDependentQs=hcAc(TcTin)hc=20W/(m2K); Ac=8m2;
Cc=1500J/(kgK)
Tc(k+1)=Tc(k)Qs/(Ccmc)
BTMS cooling/heating loadQbBTMSDependentQb=m˙Cliquid(TinTout)m˙: BTMS coolant flowrate
Tcin: temperature before heat exchanger
Tcout: temperature after heat exchanger
Physical termSymbolheat sourceTemperatureEstimation (W)Descriptions and highlights
Conduction/convection load via roof panelQcrExternal air solar radiationDependentEq. 9 in Ref. [17]
or without radiationδ: thickness; λ: conductivity; A=3.6m2;
Qcr=kcrA(TexTin)hin=25W/(m2K); hex=4.65+13.96v
kcr=(δλ+1hin+1hex)1
Conduction/convection load via windowsQcwExternal air solar radiationDependentEq. 9 in Ref. [17]
without radiationA=1.5m2; hin=25W/(m2K);
Qcw=kcwA(TexTin)hex=4.65+13.96v
kcw=(δλ+1hin+1hex)1
Solar radiation through windowsQrSolar fluxIndependentQr=i=1n=4ηIAisinθiβFour windows: windshield, rear, left, right;
β: shading factor; θ: installation angle;
η: penetration rate; I: incident radiation
Human body thermal loadQhDriver and passenger(s)IndependentQh=145+116n [40]n: passengers number; n=3
Fresh air thermal loadQfFresh airDependentQf=ξm˙airCair(TexTin)Ventilation fresh air portion ξ=12%;
m˙air=0.186kg/s
Sensible heat loadQsCabin interiorDependentQs=hcAc(TcTin)hc=20W/(m2K); Ac=8m2;
Cc=1500J/(kgK)
Tc(k+1)=Tc(k)Qs/(Ccmc)
BTMS cooling/heating loadQbBTMSDependentQb=m˙Cliquid(TinTout)m˙: BTMS coolant flowrate
Tcin: temperature before heat exchanger
Tcout: temperature after heat exchanger

3.4 Battery Thermal Management Model

3.4.1 Battery Pack Thermal Model.

As discussed in Sec. 1.1, there exist multiple thermal control technologies aiming to maintain the battery temperature within an appropriate range. Inspired by the widely used liquid cooling technology [45], we have developed a PCM-assisted plate cooling battery thermal control system, as illustrated in Fig. 7. Each battery module consists of three liquid cooling plates vertically, and in between are two battery layers. Multiple modules can be arranged to form a whole battery pack in the horizontal direction. The flow field consists of two sets of symmetric S-shape cooling channels, and the spaces in between are filled with PCM, aiming to improve the thermal performance of the cooling plate. The predefined paraffin-based RT42 PCM material has a phase-transition temperature range of Ts (41C) to Tl (42C), meaning that the PCM is solid below Ts, liquid above Tl, and mushy (solid/phase) in between the temperature range [46]. The liquid fraction β of the PCM is defined as follows:
(15)
where Tp denotes the averaged temperature of the PCM, which can be used to estimate the status of PCM. The motivations of adding the PCM as a supplementary here are twofold: (i) PCM is used as a thermal storage buffer to prevent any potential severe thermal impacts under intense driving conditions and (ii) PCM can store a large amount of latent heat under cold temperature environment, saving considerable amount of energy from thermal preservation, especially after long time driving. The steady thermal performance, dynamic thermal response characteristic, and mechanical/chemical sensitivity of the whole active temperature range (i.e., 10C to 50C) will be further investigated separately. Given the research scope, we only consider the dynamic properties of a specific temperature range in this study.
Fig. 7
The sketch of the liquid-PCM battery thermal control structure. The whole battery pack consists of three battery modules. Each module has a sandwich-like structure, containing three cooling plates in the vertical direction and two battery layers in between. The battery layer has a size of 40 mm in thickness, consisting of five battery bricks. Each brick is 0.39 m in length and 0.26 m in width, and the spaces in between have thermal insulation materials with a thickness of 5 mm. These settings are used to simulate the heat generation and thermal performance of the pack.
Fig. 7
The sketch of the liquid-PCM battery thermal control structure. The whole battery pack consists of three battery modules. Each module has a sandwich-like structure, containing three cooling plates in the vertical direction and two battery layers in between. The battery layer has a size of 40 mm in thickness, consisting of five battery bricks. Each brick is 0.39 m in length and 0.26 m in width, and the spaces in between have thermal insulation materials with a thickness of 5 mm. These settings are used to simulate the heat generation and thermal performance of the pack.
Close modal

Multiple parameters are required to present a specific dynamic state of the battery pack system, such as the temperature of battery bricks [Tb1, Tb3, Tb5], the temperature of PCM sections [Tpcm1, Tpcm3], and the temperature of the cooling plate [Tplate], as indicated in Fig. 7. Three parameters are regarded as the system inputs, including the battery inlet temperature Tcool_btm_in, the mass flowrate m˙cool of the coolant, and the volumetric heat generation rate G˙. The system state updates by iteration following new system inputs, as illustrated in Fig. 8.

Fig. 8
The states and inputs of the liquid-based thermal system
Fig. 8
The states and inputs of the liquid-based thermal system
Close modal
To evaluate and identify the dynamic response of the system, a total of 1000 CFD simulations are performed via the commercial software ansys fluent with the kε turbulence model. The CFD model has a meshing size of 3,500,000 after mesh dependency analysis, taking approximately 25 min to simulate a 5-s transient thermal behavior on a 12-core 1.6 GHz workstation. Based on the simulated data, a feed-forward multi-input multi-output (MIMO) neural network with three hidden layers is employed here to model the dynamic response. Determined by Bayesian optimization, the numbers of each layer are set as [100, 79, 41], which yields an RMSE of 1.5% and an MAE of 1.2%. The next system state estimated via a neural network-based black-box is expressed as follows:
(16)
In addition, the BTMS outlet coolant temperature can also be estimated as another output using a similar MIMO model, given as follows:
(17)

The system state at each time-step Xk and the BTMS outlet coolant temperature Tcool_btm_out are predicted directly through the neural network models N11 and N12, respectively. These models incorporate inputs such as battery inlet temperature, coolant mass flowrate, and heat generation rate to accurately simulate dynamic responses.

3.4.2 Battery Thermal Control Model.

As illustrated in Fig. 5, under high temperature conditions, the chiller is the only alternative for thermal control. The chiller model is defined using the energy balance equation as follows:
(18)
where Tcool_in and Tcool_out are the inlet and outlet temperatures of the BTMS coolant, respectively. To improve the efficiency, the coolant mass flowrate m˙cool should be adaptively adjusted by controlling the speed of the coolant pump for varying operating conditions. However, controlling the coupled variables is challenging, which requires a detailed dynamic thermal-fluid model to determine the mass flowrate and the coolant temperature at the chiller outlet simultaneously. For the sake of simplification, in this study, the mass flowrate is fixed at its maximum volume (m˙=0.9kg/s) using an on–off control. The coolant temperature at the chiller outlet can then be determined using Eq. (18).
In regards to the BTMS, a thermal control strategy is proposed to bridge the gap between the real-time temperature and the predefined temperature control trajectory, defined as follows:
(19)
From the CFD simulations, it is observed that the battery temperatures between two states change when the coolant is running at a specific mass flowrate and inlet temperature, as described in Eq. (16). The relationship can be extracted and expressed as follows:
(20)
In a control process, ΔTbat_dif is treated as an obtained value after control actions, while ΔTgap is the control target. The anticipated coolant temperature at the BTMS inlet is predicted as follows:
(21)
where the multi-input single-output controller model can also be regarded as an inverse function of Eq. (16). Similar to the plant model, the neuron numbers of each layer are determined as [53, 28,12] via Bayesian optimization, yielding an RMSE of 3.4% and an MAE of 2.8%. By substituting Eq. (21) into the chiller model as depicted in Eq. (18), the cooling demand of BTMS can be determined. Additionally, the power consumption of the coolant pump can be estimated based on Eq. (22):
(22)
where ηpump and Δppre represent the pump efficiency and the pressure drop of the battery system, respectively. ρcool is the coolant density.

4 Energy Management and Case Study

4.1 Model Predictive Control-Based Energy Management for Daily Commute.

Similar to the MPC-based energy strategy investigated in our previous study [17], based on the subsystem models established earlier, an MPC strategy is developed here with a cost function as follows:
(23)
where Pcab and Pbat are the AC power consumption for the cabin thermal control system and battery thermal management system, respectively. Qcab and Qbat are the cooling capacity provided by the AC system, respectively; the capacity allocations between them can be implemented via regulating value and flow direction controls. Here, we simplify the settings and only focus on optimization at the system level. The parameter Ppumb is the power demand from the battery coolant pump, and Paux represents the power consumption of other auxiliary systems, which are relatively small compared to the AC system. Note that the coefficient set [α, β, ξ] is adaptive to different conditions, which is determined using other driving cycles. This optimization problem was solved using the particle swarm optimization algorithm. By starting with its real-time control solution, the iterative stochastic optimization process can also guarantee a better solution compared with the real-time control solution. However, due to the highly nonconvex and nonlinear characteristics, a local optimum instead of a global optimum is expected for this MPC control given a limited computational time. To ensure feasibility and stability, our MPC is designed with realistic constraints that reflect the operational limits of the vehicle’s systems. These constraints include setting upper and lower bounds on the battery SoC to prevent the battery from operating in potentially harmful conditions, and imposing thermal thresholds on battery and cabin temperatures to avoid overheating or excessive cooling. Additionally, limits on control inputs, such as power allocations to the air conditioning (AC) system and the battery thermal management system (BTMS), are implemented to prevent abrupt or excessive control actions that could destabilize the system. The cost function is meticulously crafted with tailored weights that balance the need for short-term energy efficiency with the long-term health and performance of the vehicle’s systems. Despite the use of simplified models to represent the vehicle’s nonlinear dynamics, this approach ensures that our MPC framework remains practical to implement and robust across a variety of operating conditions in real-world scenarios.We acknowledge that the absence of a terminal cost is a limitation and plan to explore its inclusion in future research to further enhance the stability and performance of the system.

Based on the 5–15 s ahead velocity forecasting for a daily commute route, the initial system parameters are set with an exterior temperature of 310.15 K and a battery SoC of 0.95. The cabin temperature is targeted at 294.15 K, while the battery control temperature is aimed at 313.15 K. The upper bound of the battery is set as 317.15 K, to allow for a large margin of thermal impacts considering the usage of PCM.

The simulation results of the real-time control and the MPC-based approach are presented in Figs. 9 and 10, respectively. The PCM remains in fluid status during the driving cycles, since the temperature has been well constrained around the target value. The results show that the final SoC decreased from 0.95 to 0.8818 with MPC, while the real-time control yields an SoC of 0.8806. Compared with the real-time control, only a limited improvement is obtained via MPC for this specific driving cycle, i.e., less than 2% regarding the energy efficiency calculated by the SoC. The results are also verified based on real velocity data rather than forecasting. It is known that the MPC strategy is less effective for a steady driving stage with cruising speed, but performs well for varying conditions. Given this consideration, four potential reasons may account for this observation: (i) Compared to the hybrid driving cycles that consist of the Urban Dynamometer Driving Schedule (UDDS), Worldwide Harmonized Light Vehicles Test Cycle (WLTC), and the Highway Fuel Economy Driving Schedule (HWFET) cycles, these Dallas driving cycles are considerably smoother, with fewer intersections and complete stops. (ii) The waste heat dissipation from the battery system relies on the AC system, and when combined with the cooling demand from the cabin, the AC system runs at a high load ratio. The AC load is also very close to the regenerative power, leaving limited space for load shifting. (iii) The control intervals for AC power and BTMS power allocations are set at 5 s, making the AC system less responsive to varying power-train demands, and so does the BTMS coolant pump with on–off control. (iv) Given the time required for data telemetry and velocity forecasting in real practice, the computational time for solving the MPC problem is capped at 4 s. The solutions are local optima rather than global optima for a nonlinear nonconvex problem due to the very limited computational time.

Fig. 9
The performance of real-time control-based energy management
Fig. 9
The performance of real-time control-based energy management
Close modal
Fig. 10
The performance of MPC-based energy management
Fig. 10
The performance of MPC-based energy management
Close modal

4.2 Real-Time Energy Management Using Traffic Light Detection.

As discussed earlier, the MPC-based energy management has its drawbacks during steady driving conditions, especially when using nonlinear and nonconvex models for complex systems. It is also observed that the load shifting mainly occurs under changing conditions, i.e., the deceleration and reacceleration processes at intersections. Given these considerations, we aim to develop a real-time energy control framework to avoid the overlapping among peaks and to reuse the regenerative power instead of recharging back to the battery pack. For instance, if a deceleration is anticipated, the system is motivated to decrease the power demand of the AC system by defining a higher target temperature beforehand, allowing the AC system to be powered by regenerative energy during the deceleration. On the other hand, when a reacceleration is predicted, the system tends to lower down the manipulated temperature to prevent overlaps among different loads.

The image-based traffic light detection method developed in Sec. 2, in conjunction with real-time acceleration signals, is utilized as a mode indicator to update control parameter settings. At an intersection, the vehicle is able to activate the low-demand mode based on its location. Given the largest detection threshold of 100 m in our model, the distance for low-demand mode activation is set as 250 m, which is approximately 12–13 s prior to an intersection at a cruising speed. It is worth noting that accurate traffic light recognition can be achieved as far as 130–150 m away as reported in the literature [35], which is about 7–8 s ahead for a cruising scenario and more than 20 s ahead for a completely stop scenario.

Energy management based on traffic light detection

Algorithm 2

Reduce AC and BTMS power before intersections

If Traffic light signal is GREEN then Set AC and Battery to normal mode

Else If Traffic light signal is RED or YELLOW then

   Increase AC and BTMS power to use regenerative energy

   If Traffic light signal turns GREEN

    Reduce AC and BTMS power to avoid power demand overlapping during reacceleration

   End If

Else Maintain normal mode

End If

Adjust energy settings based on updated system status

As described in the pseudo-code of Algorithm 2, based on the detected traffic light signals, two major scenarios are predefined with a sequence of energy allocation actions: (i) For a green light signal, the vehicle switches back to the normal control mode and deactivates the low-demand mode. (ii) For a red or yellow light signal, the AC system prioritizes the use of regenerative power as much as possible during deceleration, instead of recharging the battery system. The system also determines an average AC power demand based on the estimated waiting time at the intersection. The AC system operates at a relatively higher power level to cool down the vehicle when a vehicle stops before the intersection, avoiding power demand overlap during reacceleration. The system switches back to the normal control mode upon leaving the intersection, as indicated in Fig. 11. This approach relies on the recognitions of traffic light signals, incorrect recognition may bring down the overall energy efficiency, but can not jeopardize the safety of battery system by setting up prioritized temperature thresholds. It is worth noting that the traffic light can act as an early termination signal, i.e., a green light detected in front of an intersection suggests switching back to normal control directly.

Fig. 11
The sketch of a traffic light detection-based energy management strategy at intersections (N/A indicates no image is captured or no traffic light signal is identified)
Fig. 11
The sketch of a traffic light detection-based energy management strategy at intersections (N/A indicates no image is captured or no traffic light signal is identified)
Close modal

In this study, the AC energy consumed by the battery thermal system remains unchanged due to its limited total volume. Instead, the AC energy for the cabin is adjusted based on the aforementioned principle. The simulation results of the modified control are presented in Fig. 12. As compared to real-time energy distribution, the main differences are observed at intersections. As a trade-off, the final stage SoC is improved by approximately 2.5% to 0.8823, at the cost of introducing more fluctuations in the cabin temperature. In comparison to the MPC-based energy management, the traffic light detection-based energy management has very similar performance for a smooth driving cycle. Given its computational efficiency, it is anticipated that the traffic light detection-based method could be a potential alternative for vehicle energy management. Our simulations were conducted in matlab, utilizing validated models of vehicle dynamics, battery systems, and thermal management, which are based on real-world parameters. The traffic light detection algorithm and model predictive control strategy were evaluated against realistic urban commuting scenarios to ensure the results closely mimic actual driving conditions.

Fig. 12
The performance of traffic light detection-based energy management
Fig. 12
The performance of traffic light detection-based energy management
Close modal

Despite the promising simulated results, it is important to note that these simulations do not include tests with real vehicles or hardware-in-the-loop simulations. Future work will focus on validating our approaches under actual operating conditions to confirm their effectiveness and practical applicability.

5 Conclusion

This study contributes to the development of efficient and accurate energy management solutions for electric vehicles. The proposed YOLO V2 framework for traffic light detection achieved high accuracy in detecting traffic lights, and the probability-based hybrid model showed promising results in short-term velocity forecasting, particularly for 5 s ahead. Moreover, the MPC-based energy management approach optimized the energy consumption of the battery thermal management system, while the traffic light detection-based approach improved the final stage SoC by approximately 2.5% to 0.8823, with computational efficiency. These findings highlight the effectiveness of the proposed methods in improving energy efficiency and reducing emissions in urban commuting routes. The results indicate that using traffic light detection for real-time control could be a viable alternative for managing energy consumption in urban commuting routes.

While the MPC-based approach was used to optimize the energy consumption of the battery thermal management system, it served primarily as a benchmark in this study. We recognize that NMPC applications in real-world scenarios may yield only local optima due to computational constraints. However, our innovative approach of integrating real-time traffic light detection for energy management is both practical and efficient. This method is currently underutilized in the automotive industry, demonstrating its potential for significant impact and further exploration.

Potential future work includes (i) enhancing the localized model selection and averaging framework by collecting more data and investigating reinforcement learning for model selection and (ii) identifying and detecting traffic signals through CNN-based image identification techniques to further improve the waiting time estimation.

Footnotes

Conflict of Interest

There are no conflicts of interest.

Data Availability Statement

The data and information that support the findings of this article are freely available.3

References

1.
Un-Noor
,
F.
,
Padmanaban
,
S.
,
Mihet-Popa
,
L.
,
Mollah
,
M. N.
, and
Hossain
,
E.
,
2017
, “
A Comprehensive Study of Key Electric Vehicle (EV) Components, Technologies, Challenges, Impacts, and Future Direction of Development
,”
Energies
,
10
(
8
), p.
1217
.
2.
Ling
,
Z.
,
Wang
,
F.
,
Fang
,
X.
,
Gao
,
X.
, and
Zhang
,
Z.
,
2015
, “
A Hybrid Thermal Management System for Lithium Ion Batteries Combining Phase Change Materials With Forced-Air Cooling
,”
Appl. Energy.
,
148
, pp.
403
409
.
3.
Liu
,
Y.
, and
Zhang
,
J.
,
2019
, “
Design a J-Type Air-Based Battery Thermal Management System Through Surrogate-Based Optimization
,”
Appl. Energy.
,
252
, p.
113426
.
4.
Zhuang
,
W.
,
Liu
,
Z.
,
Su
,
H.
, and
Chen
,
G.
,
2021
, “
An Intelligent Thermal Management System for Optimized Lithium-Ion Battery Pack
,”
Appl. Therm. Eng.
,
189
, p.
116767
.
5.
Widyantara
,
R. D.
,
Naufal
,
M. A.
,
Sambegoro
,
P. L.
,
Nurprasetio
,
I. P.
,
Triawan
,
F.
,
Djamari
,
D. W.
,
Nandiyanto
,
A. B. D.
,
Budiman
,
B. A.
, and
Aziz
,
M.
,
2021
, “
Low-Cost Air-Cooling System Optimization on Battery Pack of Electric Vehicle
,”
Energies
,
14
(
23
), p.
7954
.
6.
Wang
,
X.
,
Liu
,
Y.
,
Sun
,
W.
,
Song
,
X.
, and
Zhang
,
J.
,
2018
, “
Multidisciplinary and Multifidelity Design Optimization of Electric Vehicle Battery Thermal Management System
,”
ASME J. Mech. Des.
,
140
(
9
), p.
094501
.
7.
Amini
,
M. R.
,
Wang
,
H.
,
Gong
,
X.
,
Liao-McPherson
,
D.
,
Kolmanovsky
,
I.
, and
Sun
,
J.
,
2019
, “
Cabin and Battery Thermal Management of Connected and Automated HEVs for Improved Energy Efficiency Using Hierarchical Model Predictive Control
,”
IEEE Trans. Control Syst. Technol.
,
28
(
5
), pp.
1711
1726
.
8.
Tang
,
X.
,
Guo
,
Q.
,
Li
,
M.
,
Wei
,
C.
,
Pan
,
Z.
, and
Wang
,
Y.
,
2021
, “
Performance Analysis on Liquid-Cooled Battery Thermal Management for Electric Vehicles Based on Machine Learning
,”
J. Power. Sources.
,
494
, p.
229727
.
9.
Park
,
S.
, and
Ahn
,
C.
,
2021
, “
Model Predictive Control With Stochastically Approximated Cost-to-Go for Battery Cooling System of Electric Vehicles
,”
IEEE Trans. Vehicular Technol.
,
70
(
5
), pp.
4312
4323
.
10.
Zhao
,
S.
, and
Mi
,
C. C.
,
2021
, “
A Two-Stage Real-time Optimized EV Battery Cooling Control Based on Hierarchical and Iterative Dynamic Programming and MPC
,”
IEEE Trans. Intell. Transp. Syst.
,
23
(
8
), pp.
11677
11687
.
11.
Zhang
,
X.
,
Kong
,
X.
,
Li
,
G.
, and
Li
,
J.
,
2014
, “
Thermodynamic Assessment of Active Cooling/Heating Methods for Lithium-Ion Batteries of Electric Vehicles in Extreme Conditions
,”
Energy
,
64
, pp.
1092
1101
.
12.
Huo
,
Y.
, and
Rao
,
Z.
,
2017
, “
Investigation of Phase Change Material Based Battery Thermal Management at Cold Temperature Using Lattice Boltzmann Method
,”
Energy. Convers. Manage.
,
133
, pp.
204
215
.
13.
Yang
,
Z.
,
Mamun
,
A.-A.
,
Makam
,
S.
, and
Okma
,
C.
,
2020
, “An Empirical Aging Model for Lithium-Ion Battery and Validation Using Real-Life Driving Scenarios,” Technical Report, SAE Technical Paper, Report No. 2020-01-0449.
14.
Bi
,
Y.
,
Yin
,
Y.
, and
Choe
,
S.-Y.
,
2020
, “
Online State of Health and Aging Parameter Estimation Using a Physics-Based Life Model With a Particle Filter
,”
J. Power. Sources.
,
476
, p.
228655
.
15.
Liu
,
H.
,
Wei
,
Z.
,
He
,
W.
, and
Zhao
,
J.
,
2017
, “
Thermal Issues About Li-Ion Batteries and Recent Progress in Battery Thermal Management Systems: A Review
,”
Energy. Convers. Manage.
,
150
, pp.
304
330
.
16.
Amini
,
M. R.
,
Kolmanovsky
,
I.
, and
Sun
,
J.
,
2020
, “
Hierarchical MPC for Robust Eco-Cooling of Connected and Automated Vehicles and Its Application to Electric Vehicle Battery Thermal Management
,”
IEEE Trans. Control Syst. Technol.
,
29
(
1
), pp.
316
328
.
17.
Liu
,
Y.
, and
Zhang
,
J.
,
2021
, “
Electric Vehicle Battery Thermal and Cabin Climate Management Based on Model Predictive Control
,”
ASME J. Mech. Des.
,
143
(
3
), p.
031705
.
18.
Piao
,
X.
,
Wang
,
X.
, and
Han
,
K.
,
2021
, “
Hierarchical Model Predictive Control for Optimization of Vehicle Speed and Battery Thermal Using Vehicle Connectivity
,”
IEEE Access
,
9
, p.
141378
.
19.
Wu
,
J.
,
Wei
,
Z.
,
Liu
,
K.
,
Quan
,
Z.
, and
Li
,
Y.
,
2020
, “
Battery-Involved Energy Management for Hybrid Electric Bus Based on Expert-Assistance Deep Deterministic Policy Gradient Algorithm
,”
IEEE Trans. Vehicular Technol.
,
69
(
11
), pp.
12786
12796
.
20.
Xu
,
D.
,
Zheng
,
C.
,
Cui
,
Y.
,
Fu
,
S.
,
Kim
,
N.
, and
Cha
,
S. W.
,
2022
, “
Recent Progress in Learning Algorithms Applied in Energy Management of Hybrid Vehicles: A Comprehensive Review
,”
Int. J. Precision Eng. Manuf.-Green Technol.
,
10
(
1
), pp.
1
23
.
21.
Ehsani
,
M.
,
Singh
,
K. V.
,
Bansal
,
H. O.
, and
Mehrjardi
,
R. T.
,
2021
, “
State of the Art and Trends in Electric and Hybrid Electric Vehicles
,”
Proc. IEEE
,
109
(
6
), pp.
967
984
.
22.
Amini
,
M. R.
,
Kolmanovsky
,
I.
, and
Sun
,
J.
,
2021
, “
Hierarchical MPC for Robust Eco-Cooling of Connected and Automated Vehicles and Its Application to Electric Vehicle Battery Thermal Management
,”
IEEE Trans. Control Syst. Technol.
,
29
(
1
), pp.
316
328
.
23.
Jing
,
J.
,
Filev
,
D.
,
Kurt
,
A.
,
Özatay
,
E.
,
Michelini
,
J.
, and
Özgüner
,
Ü.
,
2017
, “
Vehicle Speed Prediction Using a Cooperative Method of Fuzzy Markov Model and Auto-Regressive Model
,”
2017 IEEE Intelligent Vehicles Symposium (IV)
,
Redondo Beach, CA
,
June 11–14
, IEEE, pp.
881
886
.
24.
Zhou
,
Y.
,
Ravey
,
A.
, and
Pera
,
M.-C.
,
2019
, “
A Velocity Prediction Method Based on Self-learning Multi-step Markov Chain
,”
IECON 2019 45th Annual Conference of the IEEE Industrial Electronics Society
,
Lisbon, Portugal
,
Oct. 14–17
,
IEEE
, Vol. 1, pp.
2598
2603
.
25.
Liu
,
Y.
, and
Zhang
,
J.
,
2021
, “
A Repeated Commuting Driving Cycle Dataset With Application to Short-Term Vehicle Velocity Forecasting
,”
J. Auton. Vehicles Syst.
,
1
(
3
), p.
031002
.
26.
Sun
,
C.
,
Hu
,
X.
,
Moura
,
S. J.
, and
Sun
,
F.
,
2014
, “
Velocity Predictors for Predictive Energy Management in Hybrid Electric Vehicles
,”
IEEE Trans. Control Syst. Technol.
,
23
(
3
), pp.
1197
1204
.
27.
Liu
,
K.
,
Asher
,
Z.
,
Gong
,
X.
,
Huang
,
M.
, and
Kolmanovsky
,
I.
,
2019
, “Vehicle Velocity Prediction and Energy Management Strategy Part 1: Deterministic and Stochastic Vehicle Velocity Prediction Using Machine Learning,” Technical Report, SAE Technical Paper, Report No. 2019-01-1051.
28.
Shen
,
H.
,
Wang
,
Z.
,
Zhou
,
X.
,
Lamantia
,
M.
,
Yang
,
K.
,
Chen
,
P.
, and
Wang
,
J.
,
2022
, “
Electric Vehicle Velocity and Energy Consumption Predictions Using Transformer and Markov-Chain Monte Carlo
,”
IEEE Trans. Transp. Electrification
,
8
(
3
), pp.
3836
3847
.
29.
Lemieux
,
J.
, and
Ma
,
Y.
,
2015
, “
Vehicle Speed Prediction Using Deep Learning
,”
2015 IEEE Vehicle Power and Propulsion Conference (VPPC)
,
Montreal, QC, Canada
,
Oct. 19–22
, pp.
1
5
.
30.
Moser
,
D.
,
Waschl
,
H.
,
Schmied
,
R.
,
Efendic
,
H.
, and
del Re
,
L.
,
2015
, “
Short Term Prediction of a Vehicle’s Velocity Trajectory Using ITS
,”
SAE Int. J. Passenger Cars-Electron. Electr. Syst.
,
8
, pp.
364
370
.
31.
Zhang
,
F.
,
Xi
,
J.
, and
Langari
,
R.
,
2016
, “
Real-Time Energy Management Strategy Based on Velocity Forecasts Using V2V and V2I Communications
,”
IEEE Trans. Intell. Transp. Syst.
,
18
(
2
), pp.
416
430
.
32.
Shin
,
J.
,
Yeon
,
K.
,
Kim
,
S.
,
Sunwoo
,
M.
, and
Han
,
M.
,
2021
, “
Comparative Study of Markov Chain With Recurrent Neural Network for Short Term Velocity Prediction Implemented on an Embedded System
,”
IEEE Access
,
9
, pp.
24755
24767
.
33.
NCTCG
, “
Government Report Transportation System Management
,” https://www.nctcog.org/trans/manage/its/transportation-systems-management, Accessed December 15, 2023.
34.
Jiao
,
L.
,
Zhang
,
F.
,
Liu
,
F.
,
Yang
,
S.
,
Li
,
L.
,
Feng
,
Z.
, and
Qu
,
R.
,
2019
, “
A Survey of Deep Learning-Based Object Detection
,”
IEEE Access
,
7
, pp.
128837
128868
.
35.
Yoneda
,
K.
,
Kuramoto
,
A.
,
Suganuma
,
N.
,
Asaka
,
T.
,
Aldibaja
,
M.
, and
Yanase
,
R.
,
2020
, “
Robust Traffic Light and Arrow Detection Using Digital Map With Spatial Prior Information for Automated Driving
,”
Sensors
,
20
(
4
), p.
1181
.
36.
Yang
,
Z.
,
Patil
,
D.
, and
Fahimi
,
B.
,
2018
, “
Online Estimation of Capacity Fade and Power Fade of Lithium-Ion Batteries Based on Input–Output Response Technique
,”
IEEE Trans. Transp. Electrification
,
4
(
1
), pp.
147
156
.
37.
Yang
,
Z.
,
Patil
,
D.
, and
Fahimi
,
B.
,
2019
, “
Electrothermal Modeling of Lithium-Ion Batteries for Electric Vehicles
,”
IEEE Trans. Vehicular Technol.
,
68
(
1
), pp.
170
179
.
38.
Huynh
,
T. A.
, and
Hsieh
,
M.-F.
,
2018
, “
Performance Analysis of Permanent Magnet Motors for Electric Vehicles (EV) Traction Considering Driving Cycles
,”
Energies
,
11
(
6
), p.
1385
.
39.
Lajunen
,
A.
,
Yang
,
Y.
, and
Emadi
,
A.
,
2020
, “
Review of Cabin Thermal Management for Electrified Passenger Vehicles
,”
IEEE Trans. Vehicular Technol.
,
69
(
6
), pp.
6025
6040
.
40.
He
,
H.
,
Jia
,
H.
,
Sun
,
C.
, and
Sun
,
F.
,
2018
, “
Stochastic Model Predictive Control of Air Conditioning System for Electric Vehicles: Sensitivity Study, Comparison, and Improvement
,”
IEEE Trans. Ind. Inform.
,
14
(
9
), pp.
4179
4189
.
41.
Pino
,
F. J.
,
Marcos
,
D.
,
Bordons
,
C.
, and
Rosa
,
F.
,
2015
, “
Car Air-Conditioning Considerations on Hydrogen Consumption in Fuel Cell and Driving Limitations
,”
Int. J. Hydrogen. Energy.
,
40
(
35
), pp.
11696
11703
.
42.
Kiss
,
T.
,
Chaney
,
L.
, and
Meyer
,
J.
,
2013
, “New Automotive Air Conditioning System Simulation Tool Developed in MATLAB/Simulink,”
National Renewable Energy Lab (NREL)
,
Golden, CO
, Technical Report, Report No. 2013-01-0850.
43.
Zhang
,
Z.
,
Li
,
W.
,
Shi
,
J.
, and
Chen
,
J.
,
2016
, “
A Study on Electric Vehicle Heat Pump Systems in Cold Climates
,”
Energies
,
9
(
11
), p.
881
.
44.
Shen
,
M.
, and
Gao
,
Q.
,
2020
, “
System Simulation on Refrigerant-Based Battery Thermal Management Technology for Electric Vehicles
,”
Energy. Convers. Manage.
,
203
, p.
112176
.
45.
Rao
,
Z.
,
Wang
,
Q.
, and
Huang
,
C.
,
2016
, “
Investigation of the Thermal Performance of Phase Change Material/Mini-Channel Coupled Battery Thermal Management System
,”
Appl. Energy.
,
164
, pp.
659
669
.
46.
Sharma
,
A.
,
Tyagi
,
V. V.
,
Chen
,
C.
, and
Buddhi
,
D.
,
2009
, “
Review on Thermal Energy Storage With Phase Change Materials and Applications
,”
Renewable. Sustainable. Energy. Rev.
,
13
(
2
), pp.
318
345
.