永磁球形电机的神经网络自适应控制算法

介绍了具有三维位置检测功能的三自由度永磁球形电机工作原理及其运动控制方法.结合球电机结构特性，基于XYZ 欧拉角和拉格朗日方程建立了球电机运动学及动力学模型.为了解决动力学模型建模误差、摩擦力和外界干扰等对球电机工作性能的影响，设计了神经网络逼近模型，并将其应用于球电机的控制之中.仿真结果表明这种控制方法与传统方法相比可以有效提高球电机的抗干扰能力，在建模误差比较大的情况下仍可以保证球电机的运动精度，其结果可以为进一步的球电机精密运动控制理论研究和实验方案设计提供借鉴和参考.

Neural network adaptive control algorithm for permanent magnet spherical motor

A control method for a 3-DOF spherical motor was introduced. The 3-DOF spherical motor had three-dimensional orientation measurement ability. Considering the mechanical structure of the spherical motor, XYZ Euler angle was used for the construction of the kinematics model. The dynamic model was established by the Lagrange's equation. Because of the dynamic modeling error, friction and external disturbances, the approximation neural network system model was applied to the spherical motor's control. The simulation results show that comparing to the traditional method, the proposed method can significantly improve the motion performance. The results can serve as a basis for the future precision motion control research and experiment.