Temperature Model Prediction of Thermostatic Water Bath Based on Long Short-term Memory Network
[Purposes]Since the temperature system of thermostatic water bath has the characteristics of strong nonlinearity and large lag,this study proposes a temperature model prediction method of thermo-static water bath based on long short-term memory network.[Methods]Firstly,the collected data were stan-dardized to find the optimal structure and hyperparameters of the long short-term memory network,which were used to fit the best data mapping characteristics and construct a dynamic mathematical model of the temperature of the thermostatic water bath.Secondly,the temperature trend in the future is predicted by the model.Finally,the results predicted by the method proposed in this study are compared with those predicted by the least squares method.[Findings]The fitting degree of the model constructed by the method proposed in this study reached 98.2%,and the MSE and MAE of the prediction results were 4.616 and 0.823 lower than those of the least squares model,respectively.[Conclusions]The method proposed in this study has higher prediction accuracy,which is of great significance to improve the production efficiency and control accuracy of thermostatic water bath.
thermostatic water bathLSTMtemperature predictionmathematical model
万方数据
维普
