汽车车道保持系统的BP神经网络控制

Vehicle Lane Keeping System Control with BP Neural Network

针对汽车侧向动力学的强非线性且参数时变特性,利用神经网络对任意非线性函数的逼近能力,并结合驾驶员操纵行为特性,建立了基于BP神经网络的汽车车道保持系统的自适应PID控制算法。不同车速和路况的仿真结果表明,该控制算法可有效地控制汽车按照预期轨迹行驶,且对车速和道路曲率的变化具有较强的适应性和鲁棒性。

According to the nonlinear and parameter time-varying characteristics of vehicle lateral dynamics, a novel vehicle lane keeping system's adaptive PID control algorithm with BP neural network was proposed, based on the ability that BP neural network can approach any nonlinear function and driver handling behavior. The simulation results under different velocities and road situations show this algorithm can effectively control vehicle to run according to the pre-given trajectory, and have good robustness and adaptability for the change of velocity and road curvature.