A Vehicle-Road Cooperative Positioning Method for Intelligent and Connected Vehicles Based on Wolf Pack Algorithm Optimized Neural Network

Vehicle-road cooperative positioning is an effective method for intelligent and connected vehicles to improve positioning accuracy. In the process of vehicle-road cooperative positioning, the height difference of roadside cooperative units can lead to the amplification of Z-axis positioning errors. In this paper, a back propagation (BP) neural network optimized by an improved wolf pack algorithm (WPA) is proposed to improve Z-axis and overall positioning accuracy. By analyzing the nonlinear propagation of vertical ranging noise, the improved WPA integrates logistic chaotic mapping for global parameter initialization, adaptive dynamic step size adjustment to balance exploration and exploitation, and t- distribution mutation to escape local optima. Simulation experiments are conducted in MATLAB environment. The experimental results show that the proposed vehicle-road cooperative positioning method can effectively reduce Z-axis and overall positioning errors, improving vehicle positioning accuracy.