基于神经网络的混沌控制原理对切削颤振的模拟控制研究

Control of Chaotic Systems Based on Neural Network and its Application to Simulant Study of Turning Flutter

在机械加工过程中存在一种以残留振痕作为机械反馈量的自激振荡 ,严重影响了加工质量、车床及刀具的寿命。理论研究与仿真结果表明 ,将人工神经网络理论引入非线性或不稳定系统混沌行为的控制 ,可以有效地减少残留振痕引起的车削颤振 ,提高车削稳定性。

As a result of the mechanic feedback, turning flutter resulted from former vibration trace is a kind of self excited oscillation. In this paper a new method of chaotic control with multilayer feed forward neural network is presented. With this method the chaotic behaviors of nonlinear or unstable system are controlled, and the self excited oscillation in turning process is eliminated.