论文部分内容阅读
在工业机器人两步辨识过程中都存在结构参数辨识矩阵奇异的问题。在转动参数辨识过程中通过编码器获得与辨识矩阵奇异列相对应的结构参数,利用部分矩阵处理法避免了转动参数辨识过程中的奇异问题;在其余参数辨识过程中,通过提前测得机器人某一特征参数,解决了由于辨识矩阵本身特点引起的参数不可辨识问题。在解决结构参数辨识奇异性问题基础之上,可以准确得到工业机器人结构参数的线性解。转动结构参数辨识过程中引入牛顿迭代法,辨识结果精度提高,其余参数辨识过程中通过分离与结构参数无关项,得到关于结构参数的线性方程,从而得到其线性解。
In the two-step identification of industrial robots, there exists the singularity of the structural parameter identification matrix. In the rotation parameter identification process, the structural parameters corresponding to the singular columns of the identification matrix are obtained by the encoder. The partial matrix processing method is used to avoid the singularity problem in the rotation parameter identification process. In the other parameters identification process, A characteristic parameter to solve the problem of unidentifiable parameters due to the characteristics of the identification matrix itself. On the basis of solving the singularity problem of structural parameter identification, the linear solution of structural parameters of industrial robots can be obtained accurately. The Newton iteration method is introduced in the identification of rotating structure parameters, and the accuracy of identification results is improved. During the identification of the other parameters, the linear equations about the structural parameters are obtained by separating the parameters which have nothing to do with the structural parameters to obtain the linear solution.