Study On Forecasting Method For Low Voltage Area Load Based On On Swarm Intelligence

Low load forecasting is to achieve low-voltage power line loss the only way of fine management. Through accurate forecasting, it is reasonable, economical low voltage operation mode of adjustment to reduce the reserve capacity of higher power, and low voltage distribution transformer load rate adjustment within reasonable limits. Meanwhile, power companies in terms of daily work, you can arrange maintenance program to reduce operating costs and increase economic efficiency. Therefore, the low-voltage power load forecasting is extremely important practical significance. Low-voltage power grid electricity consumption as there is a big customer of arbitrariness, the domestic load current forecast of low voltage areas, relies mainly on the charge of operating personnel or high voltage level to determine the load forecasting methods simple and portable, cannot meet the practical requirements for power production.This area of low voltage load forecasting the possibility of a comprehensive assessment, analysis of current shortcomings in the study. On this basis, the existing neural network to improve and build intelligent optimization methods based on multilayer feed forward neural network. According to neural network weights and threshold characteristics, building parameters genes and improved particle swarm algorithm to determine the parameters of genes. After the test, compared with the traditional forecasting methods, the improved neural network to improve the prediction accuracy, low-voltage customers can adapt the randomness and low voltage electricity load forecasting broad spectrum of areas.