论文部分内容阅读
根据大摆角五轴联动混联机床中并联模块的结构特点,提出两种求解位置正解的方法:一是以并联模块中各个驱动支链之间的距离为约束条件,提出一种解析方法求解正运动学解;二是以位置反解结果作为训练样本,提出一种利用多层前向神经网络求解机构位置正解的方法,通过构造神经网络并采用levenbergmarquardt算法训练,实现了机床并联模块从关节变量空间到工作空间的非线性映射,从而求得其运动学正解。结果表明:与解析法相比,该方法计算精度高,计算过程简洁,耗时少,可用于该机床工作空间的求解和控制。
According to the structural characteristics of the parallel module in the five-axis machine tool with large swing angle, two methods are proposed to solve the positive position. First, the distance between each driving branch in the parallel module is used as the constraint condition, and an analytic method is proposed Positive kinematic solution; the second is to use the inverse solution of the position as a training sample, a method to solve the mechanism position positive solution using multi-layer forward neural network is proposed. By constructing the neural network and training with levenbergmarquardt algorithm, Variable space to the workspace of the non-linear mapping, so as to obtain the kinematic positive solution. The results show that compared with the analytical method, this method has high computational accuracy, simple calculation process and less time consumption, which can be used to solve and control the machine tool workspace.