基于BP神经网络的排爆机械臂逆运动学分析

机械臂逆运动学是已知末端执行器的位姿求解机械臂各关节变量,主要用于机械臂末端执行器的精确定位和轨迹规划,如何高效的求解机械臂运动学逆解是机械臂轨迹控制的难点。针对传统的机械臂逆运动学求解方法复杂且存在多解等问题,提出一种基于BP神经网络的机械臂逆运动学求解方法。以四自由度机械臂为研究对象,对其运动学原理进行分析,建立BP神经网络模型并对神经网络算法进行改进,最后使用MATLAB进行仿真验证。仿真结果表明：使用BP神经网络模型求解机械臂逆运动学问题设计过程简单,求解精度较高,一定程度上避免了传统方法的不足,是一种可行的机械臂逆运动学求解方法。

Inverse Kinematics Analysis of Anti-bomb Robot Based on BP Neural Network

Manipulator inverse trajectory which means to solve the joint variable by the end of the mechanical arm position is mainly used in the accurate location and trajectory planning of the end executor. How to efficiently solve the inverse solution of kinematics of robotic is the difficulty. In view of the traditional algorithm of robot arm inverse kinematics is complex and there are multiple solutions, this paper proposes a mechanical arm inverse kinematics method based on BP neural network. With four degrees of freedom mechanical arm as the research object, analyses its kinematics principle, on the basis of improving the BP neural network algorithm, the mechanical arm are simulated and modeled by using MATLAB toolbox. The results show that using BP neural network model to solve inverse kinematics which avoids the disadvantages of traditional methods is simple in design and precision; it is a feasible manipulator inverse kinematics solution.