Studying criteria ensuring better implementations of control laws and logics on digital programmable devices is today a crucial point in order to improve performance and safety in nuclear plants of novel design. To this aim, a mathematical model of a PWR is considered, describing the dynamics of the primary circuit. This model is simple enough to allow the determination of a control policy, but it is also sufficiently accurate to capture the nonlinear, time–varying, and switching nature of the system. A dynamic controller for the pressurizer pressure and water level is hence designed. Its main advantage is the reduction of the control response. This latter is a classical drawback, which may cause some troubles to pressure control, due to the long response of temperature sensors.
The digital implementation of the controller has been taken into account to study the important aspects of the influence of the implementation on the performance of the control law. A particularly effective and flexible environment where it is possible to analyze these aspects is Matlab/Simulink, a general–purpose tool for analysis and simulation of multi–domain dynamic systems, practically a standard in control implementations, also in industrial applications. The designed controller ensures a good performance when applied to the model used to derive them, also in the presence of unmodeled uncertainties and disturbances. Its nonlinear switching nature, reflecting the real pressurizer dynamics, ensures better transient behaviors. Since it contains a PI action, it represents an evolution and an improvement with respect to classical PID controllers, usually implemented in standard control actions. To better analyze the control performance, three steps have been considered: first, the designed nonlinear controller has been compared with a linear one. Second, it has been tested on a more detailed model, with more realistic dynamics. Finally, the digital implementation of the controller has been simulated, in order to optimize the electronic implementation on the FPGA–CPU hybrid platform.
A simulation study has been conducted with sampling times in ascending order, to further test its robustness against delays. LabVIEW development tool and technology has been chosen to create the electronic layout and to obtain preliminary results in terms of timing, resources availability and power consumption. Due to the same graphical mode of programming, an immediate link between Matlab/Scilab and LabVIEW is possible in order to optimize the use of each language where it guarantees the best performance.
The methodology here presented could be very useful to better integrate the different features of control engineering and IC design, thus giving the possibility of realizing more accurate simulations and permitting to perform a hardware in the loop (HIL) testing of a wide variety of control algorithms.