基于遗传算法的电动助力转向系统鲁棒H_∞控制

Robust H-infinity control for electric power steering based on genetic algorithm

针对电动助力转向系统(EPS)中存在的模型不确定性和路面干扰问题,提出了基于遗传算法的鲁棒H∞控制方法.构建了EPS系统数学模型,以驾驶员获得较小的干扰波动和卓越的鲁棒性为控制目标,运用鲁棒H∞方法极小化干扰问题,将系统设计中加权函数的选取表示成多目标问题,用遗传算法对其优化求解,得到鲁棒控制器.分析了受到路面干扰时,方向盘把持转矩的响应情况.仿真结果表明,遗传优化后的EPS鲁棒控制器有效地增强了系统的鲁棒稳定性,提高了系统的抗干扰能力,使驾驶员获得满意的路感,提高了行驶安全性.

To deal with the model uncertainties and road disturbances in the electric power steering（EPS） system,we propose a robust H-infinity control scheme based on genetic algorithm。In this scheme,the model of the EPS system is built and the robust H-infinity method is applied to minimize the impact of road disturbances to the driver’s steering wheel for obtaining the maximal robustness.The weighted matrices in the design are obtained by optimizing multiple objectives using the genetic algorithm.The torque on the steering wheel produced by the road disturbance is analyzed.Simulation results show that this robust controller enhances the robustness and the ability of disturbance rejection,consequently,improving the driving safety and comfort for the driver.