基于朴素贝叶斯算法的变权重多目标自适应巡航控制策略

Strategy of multi-objective adaptive cruise control system with variable weight based on naive Bayesian algorithm

针对目前使用较为广泛的基于模型预测控制的自适应巡航系统,该文提出一种分层控制结构,设计一种变权重的模型预测控制器作为上位控制器。为了获得最佳加速度,综合考虑驾驶员期望车距,车辆自身物理限制,前车加速度影响等因素。使用高斯朴素贝叶斯算法预测前方车辆未来行为,从而采取不同的权重参数策略。通过Matlab/Simulink与Carsim对固定权重参数和可变权重参数分别进行联合仿真,结果表明,随着道路条件的变化,可变权重参数可以提高车辆自适应巡航系统的表现效果,显著降低固定权重策略的系统速度和距离偏差量,有效提高了系统的控制精度与适应性。

Aiming at the widely used adaptive cruise control system based on model predictive control,a hierarchical control structure is proposed,which is divided into upper controller and lower controller.A model predictive controller with variable weight is designed as the upper controller.In order to obtain the best acceleration,the expected distance of the driver,the physical limitation of the vehicle itself,the influence of the acceleration of the vehicle in front and other factors are taken into comprehensive consideration.The weight changes according to the future lane changing behavior of vehicles in front of ego vehicle predicted by Gaussian naive Bayesian algorithm.Finally,the fixed weight parameters and variable weight parameters are simulated by Matlab/Simulink and Carsim respectively.The simulation results show that with the change of road conditions,the variable weight parameter can improve the performance of the vehicle adaptive cruise system,significantly reduce the system speed and distance deviation of the fixed weight strategy and effectively improve the control precision and adaptability of the system is improveded effectively.