基于粒子群优化算法的城市日用水量预测模型  

Forecast Model of Urban Daily Water Consumption Based on Particle Swarm Optimization

作　　者：王亮[1] 汪震[2] 岳琳[3] 

机构地区：[1]天津工业大学材料科学与化工学院,天津300160 [2]天津国际工程咨询公司,天津300203 [3]天津大学环境科学与工程学院,天津300072

出　　处：《中国给水排水》2007年第7期89-93,共5页China Water & Wastewater

基　　金：国家自然科学基金资助项目(50278062;50578108)

摘　　要：结合城市日用水量影响因素的特点和变化规律,建立了城市日用水量预测模型,采用粒子群优化算法优化BP人工神经网络的连接权值,以求解该预测模型。经优化后的BP人工神经网络运算速度快、泛化能力强、预测精度高。实例验证结果证明该日用水量预测模型和求解方法是可行的。Combined with the characteristics and variation rule of the factors influencing municipal daily water consumption, a forecast model for municipal daily water consumption was set up, then the particle swarm optimization was used to optimize BP artificial neural network to solve the model. The optimized BP artificial neural network model has advantageous properties such as more rapid calculation, high generalization performance and high accuracy. The experimental results prove the forecast model for daily water consumption and the solution are feasible.

关 键 词：城市日用水量 预测模型 粒子群优化算法 BP人工神经网络 泛化能力 

分 类 号：TU991.31[建筑科学—市政工程]