Risk Analysis Model for Water Pipeline Leakage Based on FAHP and BPNN

Scientific analysis of the leakage of the water distribution system in city is very helpful to water supply network's maintenance and renovation, and hence reduces negative social effect and economic loss. A leakage risk analysis model for water distribution system was established based on fuzzy analytical hierarchy process (FAHP) and BP neural network (BPNN). This model introduces FAHP to reasonably ensure initial state of BP neural network, and uses weighted superposition to mend learning sample set of BP neural network. The water distribution system of a city in Zhejiang province P.R.China was selected to test the proposed risk analysis model, which verifise its feasibility and effectivity.