工业机器人误差建模与仿真研究

误差模型是机器人标定的基础，针对机器人姿态的视觉测量方式，建立了适用于机器人全标定的运动学模型，在此基础上推导了以世界坐标系为标定的基准坐标系且包含所有参数的误差模型。明确指出了误差模型的参考坐标系，解决了仿真上的关键技术，研究了验证误差模型正确性和精度的公式，并对误差模型的正确性和误差辨识精度进行了仿真对比，得到了误差模型的重要特性：误差模型有一定的精度限制；随着测量次数的增加，误差辨识的精度会提高，但当测量次数达到一定数量时，辨识精度提高不明显。

Research on Industrial Robot Error Modeling and Simulation

Error model is the base of robot calibration. A kinematic model is built which is suitable for robot calibration completely according to robot vision based poses measure method, and an error model based on world coordinates and including all parameters was deduced based on it. The error model reference coordinates was pointed out clearly, and the key technology of simulation was solved. The equations for testing validation and convergence of error model was studied, and the simulations have done by comparisons methods, and the characteristics of the error model was obtained: the accuracy of the error model is limited; as the measure times were improved, the accuracy of identification would be improved, but when the measure times were arrived at some times, the improve of the accuracy of identification was not distinctness.