120 Study on Multi-Sensor Data Fusion Technology for Automotive Driving System

The purpose of the automotive driving system is to understand the surroundings of the vehicle, evaluate potential risks, and help the driver to behave correctly, safely and efficiently. And for the purpose, we need the precondition that can accurately and precisely estimate the vehicle state. One key challenge is to design a data fusion method to fuse information from multiple sensors used in the automotive driving system. This paper introduces the automotive driving system with Robocar and also presents the method of adaptive particle filter to fuse different data from different sensors to estimate the vehicle state even in noise environment. We are trying to fuse data from different sensors in a probabilistic way. A particle filter with joint observation model will be adopted to fuse different observations according to accuracy and reliability of the corresponding sensor.