摘要
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 system shows significantly multi-input/multi-output and nonlinear behaviors when turning the workpieces with complex profile. To this nonlinear system, traditional error compensation method for linear system is not valid and can not assure the required accuracy. Hence, an artificial neural network (ANN) method for high shape accuracy control is proposed. A multilayered neural network is trained off-line to identify the inverse model of the machining system. Then, the inverse model is employed to tune the system. As a result, the output is obtained as close as possible to the desired input and the machining error is minimized. This inverse model can be identified continuously, so the steady-state and the dynamic characteristics of the machining system can be compensated successfully. It is shown, by experiments of turning a piston with middle-convex shape and variable ellipticity section, that this control scheme can improve the machining accuracy effectively.
出处
《中国机械工程》
CAS
CSCD
北大核心
1996年第6期21-23,共3页
China Mechanical Engineering
基金
国家自然科学基金
