论文部分内容阅读
根据空间三自由度并联机构和X-Y二自由度运动平台组成的新型混联机床构型的结构特点,提出两种求解位置正解的方法:一是以刀具所在动平台三个球铰之间的距离为约束条件,提出一种解析方法解出位置正解;二是提出多层前向神经网络求解机构位置正解的方法,此法将位置逆解结果作为训练样本,采用levenberg-marquardt算法,实现了机构从关节变量空间到工作空间的非线性映射,从而求得位置正解。结果表明:与第一种方法相比,神经网络方法的计算精度高,没有复杂公式推导和大量的计算,计算过程简洁,耗时少,可应用于该机床的工作空间控制或是求解机床的工作空间。
According to the structural characteristics of the new hybrid machine tool consisting of space three-DOF parallel mechanism and XY two-degree-of-freedom motion platform, two methods are proposed to solve the positive position. First, the distance between the three ball joints of the moving platform As constraint conditions, an analytical solution is proposed to solve the positive position solution. Second, a multi-layer forward neural network is proposed to solve the mechanism position positive solution. This method takes the inverse position solution as a training sample and adopts levenberg-marquardt algorithm to realize the mechanism From the joint variable space to the workspace non-linear mapping, and thus find the location of positive solutions. The results show that compared with the first method, the neural network method has the advantages of high computational precision, no complex formula derivation and a large number of computations. The calculation process is simple and time-consuming, and can be applied to the workspace control or machine tool of the machine tool Workspace.