一种高速数控机床电主轴表面温度智能预测方法

A Intelligent Prediction Method for Surface Temperature of High-speed CNC Machine Tool Spindle

电主轴是高速数控机床核心功能部件,电主轴损坏基本是电主轴发热引起的。电主轴温度场具有复杂的非线性特征,神经网络在处理非线性系统温度预测方面得到了广泛的研究,神经网络与传统模型相比具有更好的适时预报性和持久性。论文利用遗传算法优化BP神经网络建立电主轴表面温度预测模型。预测结果表明,未优化的BP神经网络与遗传神经网络预测误差相对比,遗传神经网络对电主轴表面温度预测具有更高的预测精度和稳定性。

Motorized spindle is the key component of the high-speed CNC machine tools. And processing engineering motorized spindle damage is often caused by motorized spindle fever. Motorized spindle temperature field with complex nonlinear characteristics, neural net- work in dealing with nonlinear system temperature prediction has been widely research, neural network is compared with the traditional model has better predict timely and durability. In this paper, temperature prediction model of spindle surface is built based on BP artificial neural network optimized with genetic algorithm, results show that BP artificial neural network with and without genetic algorithm optimiza- tion compared the prediction result, that is to say prediction based on BP artificial neural network optimized with genetic algorithm is of higher precision and stability.