基于灰色神经网络的机床热误差建模

Grey Neural Network Modeling for Machine Tool Thermal Error

结合灰色模型和神经网络对数据处理的优点,提出了并联和嵌入型2种结构的灰色神经网络机床热误差预测模型。前者是在灰色模型和神经网络分别对机床热误差进行预测的基础上,采用线性组合方式,按照目标预测精度调整模型的加权系数,从而得到最终组合预测结果;后者是在神经网络输入层前增加灰化层,在输出层后增加白化层,通过对神经网络拓扑结构的改进,达到弱化原始数据随机性、提高预测模型鲁棒性和容错能力的目标。通过与传统灰色模型和神经网络进行试验结果对比表明:上述2种结构的灰色神经网络模型均提高了预测精度,且具有对原始数据要求低、计算简便、鲁棒性强等优点,可用于复杂实际加工场合中的数控机床热误差实时补偿。

This paper proposed a new model of prediction on thermal error of machine tools based on grey neural network combining the data processing merits of grey model and artificial neural network,respectively.The new model can be classified into two forms——parallel grey neural network（PGNN） and inlaid grey neural network（IGNN）.The former is to predict the thermal error with optimal linear combination of the result from grey model and artificial neural network respectively,while the weight value of this model is subject to the required accuracy of the experiment.The latter is to optimize the topological structure of the neural network by adding a grey layer before the input layer and a white layer after the output layer,so as to reduce the randomness of the original data and enhance the robustness and the fault-tolerant ability.Compared with the traditional grey model and the artificial neural model,the two forms of grey neural network model prove better in terms of prediction accuracy,calculation convenience and robustness.What＇s more,they require less to the original data.Thus,the new proposed models are recommended to be applied to different working environment to compensate the thermal error of machine tools.