【摘要】 由于机械手逆运动学问题的求解存在多解且非常复杂,以往解决机械手的逆运动学求解问题主要是通过神经网络逆模型来建立机械手的逆运动学模型然后通过遗传或改进的BP算法来训练神经网络的权(阀)值矩阵从而达到问题的求解,然而这种方法在建立神经网络的逆模型时要对训练数据进行限制或筛选使其成为单解问题(即满足逆映射关系存在的要求),这对于那些对数据事先进行处理很困难或根本无法进行的复杂系统是不可行,为此提出了一种采用小脑神经网络和约束条件相结合的方法来解决逆运动学问题。研究结果表明此方法可以很好的解决机械手的逆运动学控制问题,同时该方法可以推广应用到那些通过数学模型求解困难或者数学模型不确知的复杂系统反求问题。更多还原

【Abstract】 Because the solving of the Inverse Kinematics of Robots is complicated,before the inverse neural network had been applied to solve the problem ,whose weights and bias was gotten through the genetic algorithm or the approved back propagated algorithm. However this method needed us to limit or reselect the training data to meet the demand of the existence of the inverse mapping. As a result it is infeasible for the complex system for which it is very difficult or impossible to provide the right training data. Therefore the new method based on CMAC neural network and constraint condition is used to solve the problem of the inverse kinematics of robots, and the results show it is good at solving this problem. Furthermore, the proposed method is widely applied to solve the Inverse Solution of the complex system whose mathematical model is too complixated to solve or unkown for us at present .更多还原