Driver Behaviour Characteristics Identification Strategies Based on the Bionic Intelligent Algorithms

摘要

The identification strategies of driver behaviour characteristics based on the bionic intelligent algorithms are introduced in this paper. First, a driver behaviour data acquisition system is established based on the dSPACE simulation platform. Then, in order to recognize the characteristics of driver behaviour, the preview optimal curvature model is chosen as the ideal driver behaviour model. This model consists of two parts as the vehicle model and the driver model. The vehicle model parameters are identified by least square method, and the driver model parameters that could represent different types of driver behaviours are identified through genetic algorithm and particle swarm optimization method. Finally, simulations are carried out to verify the identification strategies by Carsim and Matlab/Simulink. As compared the identified results with the real driver behaviour data, there is a high similarity between them, which means the driver behaviour characteristics that have been identified are very precise and reliable.