采用限定记忆区间最小二乘法辨识DC伺服电机系统非线性参数的方法

A NON-LINEAR PARAMETER IDENTIFICATION METHOD FOR SERVOMOTOR SYSTEMS BY MEANS OF LEAST MEAN SQUARE METHOD WITH MOVING RANGE

  • 摘要: 以DC伺服电机系统作为研究对象,提出了采用限定记忆最小二乘法辨识具有非线性参数的DC伺服电机系统动力学参数的新方法.(1)采用限定记忆最小二乘法,辨识出了伺服电机的非线性参数与运动状态相关的轨迹.(2)提出了以速度指令信号的周期设定限定记忆区间长度L的方案.(3)对用限定记忆最小二乘法辨识出的非线性参数列再构出非线性参数列与状态变化间的对应关系,然后对每个与状态变化相关的非线性参数列,使用最小二乘法进行二次曲线的拟合,进一步构造出非线性参数的近似模型.(4)将拟合得到的非线性参数列的模型代入到DC伺服电机的动力学方程式中,便可构成电机系统的非线性模型.采用限定记忆最小二乘法可以观测到系统的非线性参数与运动状态变化的关系,并且可以构成更精确的电机系统的非线性模型,通过对模型与电机系统之间的误差检定,确认了使用此方法建立的模型具有误差小、精度高,更接近实际系统.

     

    Abstract: In this paper the servomotor systems are studied.A new method is proposed in which the dynamic parameters of servomotor systems with their changing state are identified by means of least square method with moving range. In the proposed method the nonlinear parameters are identified by the calculation method of step-by-step identification and approach on the bsasis of theorem and methods of linear parameter estimations.The main points are summarized in the following:(1)A scheme suggests that the length of moving range is set up in accordance with a period of identification input signal.(2)After the corresponding relations between parameters and states are ientified by least mean square method with moving range,each parameter which is changed depending on states is identified for getting the models of non-linear parameters.(3)The non-linear model of a servomotor can be obtained by substituting models of nonlinear parameters. The approximate model of a non-linear servomotor is confirmed through W examination of identification errors.This approximate model has higher precision than the linear one

     

/

返回文章
返回