Application of BP Neutral Networks to Water Demand Prediction of Shenyang City Based on Principle Component Analysis

Taking the water demand data from 2000 to 2011 of Shenyang City of Liaoning Province, this paper analyzes the main factors that influences the water resource quantity based on the principle component analysis method. The results show that agricultural population, non-agricultural population, effective irrigated area, industrial added value and precipitation are the primary indexes that affect the local water demand. Taking the main indexes as input samples, the paper set up prediction model of water demand. The stimulation results are consistent with the actual value and use the model to predict water demand in 2014 and 2015. The prediction results of BP neutral networks and index quota method are compared, and it shows that the water demand prediction by BP neutral networks is better. It is suggested to put index quota method of water demand prediction as the foundation, and combine with BP neural networks prediction results to better guide regional water resources allocation.