遗传算法与LuGre摩擦模型的非线性摩擦力观测器设计

Design of Nonlinear Friction Observer via Genetic Algorithm and LuGre Friction Model

广泛存在于机械系统中的摩擦现象在系统处于低速运动时会表现出复杂的强非线性特征,将大幅度降低系统位置控制性能,因此需要将摩擦现象进行模型化分析并纳入系统特性。对常见的摩擦模型进行了介绍和分析;并以能够较为全面描述摩擦现象丰富特性的LuGre摩擦模型作为基础,讨论并分析了其数学特性;采用遗传算法对其6个未知参数进行识别,设计了一种非线性摩擦力观测器。通过仿真与试验,对遗传算法的有效性和摩擦力观测器的效果进行了验证。试验表明,提出的摩擦观测器与传统摩擦模型相比,能使系统跟随滞后量减小40%,实现系统理想的控制性能和轨迹跟踪精度。

The friction phenomenon,which is widely present in mechanical systems.When the system is in low-velocity motion,the complex and strong nonlinear characteristics are shown,which will greatly reduce the position control performance of the system.Therefore,it is necessary to be modeled and incorporated into the system characteristics for analysis.In this paper,the common friction models are introduced and analyzed.Based on the LuGre friction model which can comprehensively describe the characteristics of friction phenomena,the mathematical characteristics of LuGre friction model is discussed and analyzed.The six unknown parameters of the models are identified by genetic algorithm.At the same time a nonlinear friction observer is designed.Finally,the effectiveness of genetic algorithm and friction observer are verified by simulation and experiment,which can effectively identify the system friction.