论文部分内容阅读
在复杂曲面的切削过程中，加工系统表现出显著的多输入多输出及非线性特征。传统的误差补偿方法不能有效地保证工精度［１，２］，因此提出一种加工误差控制方法，引入神经网络对加工系统的逆模型进行辨识，运用该模型前置校正加工系统以改善加工效果。充分考虑到机械系统的非线性特征，且网络模型可连续辨识，因而系统的静态性能和动态特性均能有效补偿。在中凸变椭圆活塞裙面加工中的成功应用，证明其合理性及先进性。
Machining systems show significant multiple-input, multiple-output and nonlinear characteristics during the cutting of complex surfaces. The traditional error compensation method can not effectively guarantee the accuracy [1, 2]. Therefore, a method of machining error control is proposed. The neural network is introduced to recognize the inverse model of the machining system. The model is calibrated using the pre-alignment system to improve the machining result . Considering the nonlinear characteristics of the mechanical system, and the network model can be continuously identified, the static performance and dynamic characteristics of the system can be effectively compensated. The successful application of convex convex oval piston skirt surface processing, to prove its rationality and advanced.