Power system load forecasting by improved principal component analysis and neural network

Power system load forecasting is the basis of power system planning and operation, in order to improve the accuracy of power system load forecasting and ensure the stable operation of power system, a number of related indices must be taken into consideration in the research of power system load forecasting. The tradition principal component analysis is always used to process these indices. In allusion to the problem that information reserved by tradition principal component analysis is not comprehensive. The improved principal component analysis is used to mean processing initial data, it can effectively eliminate the improved influence of the evaluation indices dimension and order of magnitude, reflecting information on the degree of variation of each index and the mutual influence between different indicators contained in the initial data more comprehensively, reducing the evaluation indices dimension, acquiring the principal components of power system load forecasting, it can effectively reduce the input of the radial basis function network and improve the accuracy of load forecasting of power system. The effectiveness of the proposed method is validated by results of case analysis.