Investigation On The Model Of The Short-term Load Forecasting Of Electric Power Based On Neural Network

Load Forecasting of Electric Power System is foundation to realize the safety of the electricity system and to run the economic smoothly. Fast and high precision load forecast system is the guarantee of that electric power market supply and demand balance and power system operates reliably. Under the condition of electricity market, when the load forecast in balance caused by a lot of operation cost and loss of profit, the research of the advanced intelligent load forecasting method is very necessary.This paper firstly gives summary of all kinds of existing power load forecasting methods,the present research situation and the development trend. It studies on the application of the neural network in short-term load forecasting of electric power In the basis of traditional method and the SOM algorithm in rejecting the application of bad load samples; It introduces the BP neural network, the RBF neural network and Elman neural network structures, compositions and the principle of training algorithms detail , establishes the model of the short-term load forecasting of electric power based on BP neural network, the RBF neural network, and analyses and compares the different hidden units to affect the results of model prediction; It analyzes the particle swarm optimization (PSO) principle, according to BP algorithm training Elman establish load prediction model exists disadvantages such as the slow convergence speed, it studies the PSO - BP hybrid algorithm Elman neural network training to set up short-term power load forecasting model methods; It compares the simulation results of the different short-term load forecasting of electric power based on different neural network, does a detailed analysis on the simulation results which show that neural network based load forecasting method in the application of short-term power load forecasting has bigger superiority.Finally, this paper puts forward the existing problems and the subsequent research opinions in the load forecasting model according to the actual application.