Prediction Model of Water Demand in Yinchuan City Based on CIWOA-BP and Grey Confidence Interval
Water demand prediction is the foundation of scientific allocation and scheduling of water resources.In order to improve its ration-ality,an improved whale algorithm(CIWOA)optimized BP neural network combined with grey confidence interval estimation was proposed to address the fluctuation and uncertainty characteristics of water demand changes.The influence factors of water demand were screened by grey correlation analysis.Cubic chaotic mapping and whale algorithm modified by adaptive weight were input to optimize the BP model(CI-WOA-BP).Combined with grey confidence interval estimation,the combined interval prediction model was established to simulate and fore-cast the water demand of Yinchuan City.The results show that the prediction accuracy of CIWOA-BP model is better than that of conventional whale algorithm optimization BP model(WOA-BP)and genetic algorithm optimization BP model(GA-BP).The combination model of CI-WOA-BP model and grey confidence interval is superior to the Bootstrap interval estimation model,and the prediction of water demand inter-val is reasonable and reliable when the confidence is 90%.
whale algorithmBP neural networkgrey confidence intervalwater demand forecastYinchuan City
国家科技期刊平台
NSTL
万方数据
