Gas load prediction based on DE-SVM algorithm

In order to improve the prediction accuracy, in accordance with the influence factors and characteristics of gas load, a model based on SVM (Support Vector Machine) has been established. In order to optimize the behavior of SVM, DE (Differential Evolution) algorithm was introduced into classic SVM. Using the algorithm to predict a real example and compare with SVM model optimization method based on GA (Genetic Algorithm), ACO (Ant Colony Optimization) and POS (Partial Swarm Optimization) demonstrate an improvement of generalization performance.