Short-term Load Forecasting with LS-SVM Based on Improved Ant Colony Algorithm Optimization

Research of short-term load forecasting has important practical application value in the field of power network dispatching. The regession models of least squares support vector machines (LS-SVM) have been applied to load forecasting field widely, and the regression accuracy and generalization performance of the LS-SVM models depend on a proper selection of its parameters. In this paper, a new regession model of LS-SVM was put forward based on adaptive ant colony optimization (AACO). AACO was used to optimize the kernel parameter σ and regularization parameter C of LS-SVM. Results of case analysis and simulation show that the proposed forecasting model can obtain more generalized performance and better forecasting accuracy compared with the method of single SVM and BP neural networks.