Improved IGWO-LSTM Algorithm for Modeling and Analysis of Thermal Error Prediction of Motorized Spindle
Aiming at the thermal error modeling of the motorized spindle under complex operating conditions,a LSTM neural net-work parameter prediction model IGWO-LSTM based on improved gray wolf optimization(IGWO)was proposed.By improving the cal-culation method of convergence factor a of gray wolf algorithm,its optimization performance was improved.Then,an IGWO-LSTM closed-loop system consisting of the fitness function of the IGWO algorithm and the number of nodes of the LSTM implicit layer was used to train and predict the electric spindle thermal error prediction model,so as to obtain the purpose of improving accuracy of the model and avoiding getting trapped in local optima.In order to verify the advantages of the algorithm,it was compared with the LSTM model before improved.From the calculation of mean absolute error,mean absolute percentage error and root mean square error,it is found that the three indexes of the IGWO-LSTM are better than those of the LSTM model before improved,which shows that IGWO-LSTM algorithm has better thermal error prediction accuracy and global search capability.
motorized spindlethermal errorIGWO-LSTMneural network prediction model
万方数据
维普
