Prediction method of Internet of things operation index based on improved Support vector machine regression model

Abstract

In the power system, the data on the client side is a gold mine to be mined. However, the current data mining lacks a effective influence relationship model, and it is difficult to achieve accurate prediction of power system Internet of things (IoT) indicators. Therefore, in view of the prediction problem of local IoT operation indicators of smart grid customer-side metering equipment under different climatic conditions, a correlation analysis and prediction method based on an improved support vector machine regression (SVR) model is proposed. The method first performs feature selection on climate data to improve model performance while reducing computational complexity, and then uses an optimization algorithm to further optimize model performance. For feature selection, grid search algorithm and K-fold cross-validation are used for Random Forest (RE) parameter selection. For the calculated RF feature importance, a goodness-of-fit-based sequence forward selection (Sequential Forward Selection, SFS) algorithm is used for dimension selection. After feature selection, the Particle Swarm optimization (PSO) algorithm was used to optimize the Support Vector Regression model. The experimental results show that the proposed algorithm has better performance and higher computational efficiency than the traditional combination algorithm.