Evaluation of condition of urban non revenue water based on BP neural network

For the condition of urban non revenue water in our country, this paper on the base of the scientific analysis of evaluation of the condition of urban non revenue water at home and abroad provides evaluation index system, which composed of seven indexes divided into 3 groups: physical loss or real loss, apparent loss and unbilled authorized consumption, and builds an evaluation model based on BP neural network. Though non-linear mapping function, evaluation indexes are standardized, and this paper combining with examples proves that the method is simple, accurate and effective verify the effectiveness of the method.