Research on Monitoring Data Optimization Method Based on Machine Learning

In order to analyze the variation characteristics of the measured data of the monitoring system and determine the main factors affecting the accuracy of the measured data, four models in machine learning methods were optimized for parameters and compared for prediction effects. The optimal model was selected for analysis and prediction. The results showed that the support vector machine model was the optimal choice. Through the quantitative analysis of temperature strain linear regression results, it is found that the temperature change of the bridge is correlated with the zero point change of the strain reference. Embedding the predictive model algorithm into the system to remove data changes caused by temperature effects, the optimization results show that this method can effectively eliminate the impact of temperature effects on monitoring data such as strain.