Reliability-Driven Fuzzy Rule Inference for Critical Safe Driving in Vehicle Tracking Control

As intelligent vehicles are increasingly prevailing in the modern transportation system, ensuring the safe control of vehicle tracking has emerged as a focal topic. This work delves into the modeling of longitudinal tracking behavior in vehicles, with emphases on the embedding of fuzzy logic systems into tracking control. First, to enhance the accuracy and robustness of fuzzy rule extraction, the Wang-Mendel (WM) algorithm is improved based on the reliability-driven hierarchical similarity partitioning and then a robust fuzzy controller is designed. Further, to align with drivers’ psychological expectations and safety distances, both deviation and speed differences are utilized as inputs for the fuzzy controller. Next, for some critical tracking scenarios, such as the following state losing and emergency braking, an optimal safety-driven fuzzy logic tracking control method is proposed. Finally, by establishing the headway time-based critical conditions, an optimization equation is constructed to obtain the optimal solution for the vehicle acceleration. Numerical simulations and comparative experiments validate the robustness of proposed fuzzy controller, which indicates that the model is safe and reliable under different driving conditions, contributing to the traffic flow efficiency enhancement and road safety assurance.