基于遗传算法的舵机数学模型辨识研究

Research on Mathematical Model Identification of Steering Gear Based on Genetic Algorithms

针对"飞思卡尔"杯全国大学生智能车竞赛,在智能车整体控制中,对于舵机的精准控制显得十分重要;分析了舵机系统各个部分组成原理,采用实验建模方法,利用实验测得的舵机系统的阶跃响应曲线来辨识出舵机模型参数(增益k、惯性时间常数T、滞后时间L);并根据遗传算法原理,借助MATLAB软件方便的绘图功能和强大的图像处理能力,用MATLAB语言编写了M文件进行辨识仿真;仿真结果表明,辨识参数误差小,对于舵机模型是一种有效的辨识方法;实践证明了该方案的可行性,这为实现舵机实时准确控制提供了重要的理论支持。

With regard to the ＂Freescale＂ cup of the national college students＇ intelligent car competion,the accurate control of actuator become very important in the overall domination of intelligent vehicle.In this paper,by way of analysing the structure of the intelligent car＇s actuator model as well as using the genetic algorithms,we identify the parameter of the actuator model（gain k、time constant T of the inertia、pure lag time L） with the help of M file complied by MATLAB since it has the convenient drawing function and powerful image processing ability.The result shows that it is an effective method of actuator model＇s identification which has little Parameter errors.It provides important theoretical basis to realize the real-time accurate control of actuator.