Abstract: Analyzing the deficiency of the traditional BP algorithm, combining Levenbery-Marquardt optimized algorithm and a neural network forecasting method, this paper put forward a L-M optimized BP algorithm, which quickens the train, improves stability and avoids trapping into local minimum. For some area power supply load of Power Corporation in somewhere, a short term load forecast was simulated based on L-M optimized BP algorithm. Analyzing the simulation results, it shows the L-M optimized BP algorithm has better forecast precision and adaptive capacity.

Key words: Short term load forecasting,L-M optimized,BP algorithm,Forecast precision