基于遗传算法的汽车主动悬架LQG控制器的设计

Design of an Active Suspension LQG Controller Based on Genetic Algorithm

采用合适种群规模遗传算法,寻求在一定范围内满足车身垂向加速度、轮胎垂向变形量、悬架动挠度均方根值相对于被动悬架都降低的最优加权系数,设计了基于遗传算法的汽车主动悬架线性LQG控制器。Matlab/Simulink仿真实验表明:采用合适种群规模遗传算法选取最优加权系数的用时数只需常规编程的1.17%;采用最优加权系数的LQG控制器的汽车主动悬架,可以在不影响汽车操纵稳定性和悬架行程的前提下,很好地提高汽车的行驶平顺性。

An active suspension LQG controller was designed based on genetic algorithm with appropriate population size by optimal weighing parameters which had within a certain range reduced vertical acceleration and vertical tire deformation and the root mean square of suspension deflection to the passive suspension.Results of Matlab/Simulink simulation experiment show that genetic algorithm with appropriate population size spends only about 1. 17% of the time by conventional programming in selecting optimal weighting parameters,and active suspension with LQG controller with optimal weighting parameters can improve the ride comfort of the vehicle without affecting the stability of the vehicle and suspension deflection.