Calculation And Analysis Of Power Distribution Network Line Loss On RBF Neural Network

Power distribution network line loss is a major economic indicator for assessing power enterprises, and directly affects economic benefit of electric power departments, so that the reduction of power distribution network line loss rate becomes a problem required to be solved urgently by the electric power departments. In order to achieve assisting and guiding functions for line loss management of the electric power departments, power distribution line loss is required to be accurately calculated and analyzed for making reasonable power distribution line loss reduction implementary measure and line loss assessment indicators.This paper analyzes a traditional power distribution network theoretical line loss computation method, and proposes a genetic algorithm-based RBF neural network power distribution network line loss computation method aiming at the characteristics of complex structural and operating parameters, incomplete acquisition of original data of each component and part, relatively low degree of automation and the like of power distribution network lines. In addition, this paper also introduces adaptive genetic algorithm and best preservation strategy to improve genetic algorithm for allowing the neural network to have relatively high convergence rate and accuracy which more meet actual requirements.In order to verify that the computing method proposed by this paper is more suitable for power distribution network line loss computation, we conduct simulating calculation on67lines governed by Daqing Oil Field Electric Power Group, and the simulation result proofs that this computing method is high in computation accuracy, simple in network model, and very high in applicability and popularization. In addition, this paper compares the RBF neural network after genetic optimization with the traditional RBF neural network, and the BP network, and finds out that this computing method is obviously improved in computation accuracy, memory ability than other computing methods.Finally, a set of power distribution network line loss computation software is developed under the great support of the Daqing Oil Field Electric Power Group. The main functions of the software comprise user management, database management, power distribution network line loss computation, line loss analysis, report output and the like. The human-computer interface of the software is friendly, the execution efficiency is relatively high, the layers of the data structure are clear, and the data access efficiency is high. In addition, at the later stage of the development, the software is subjected to testing by field personnel, so that integrity and the reliability of the software are guaranteed.