Research on Short-Term Driver Following Habits Based on GA-BP Neural Network

The current commercial intelligent driving systems still take the optimal strategy judged by machine to be only goal. Therefore, in order improve experience of following scene, based on assumption that environmental factors remain unchanged for a short time, five important parameters affecting scene are selected through correlation analysis, and vehicle-following research is carried out. This paper adopts driver-following model Genetic Algorithm (GA)-optimized Back Propagation (BP) neural network. Based data next-generation simulation (ngsim), this selects vehicle 32 (32 represents ID ngsim project) as main study short-term habits. A BP network built using MATLAB; 60% vehicles 29 used training set, 20% verification test set. Because prediction requires high timeliness, genetic algorithm optimize initial weights network, which not accelerates convergence speed but also plays role avoiding local solution. experimental results show compared with traditional stimulus-response model, has ability more line driver’s habits terms ensuring safety.