粗糙集算法和遗传神经网络在城市日用水量预测中的应用研究-市政工程专业论文.docxVIP

唰唰唰adw 唰唰唰adw 刷刷刷吨， 川川时 ，， 喇喇喇肉 。 刚刚刚av m川n川 υ 阳州州州4 刷 刷州Y 用水量预测中的应用研究 摘要 随着城市供水系统的范围和规模的快速扩大，供水系统日趋复杂， 传统的经验调度方式面临挑战，对供水系统实行优化调度势在必行。城 市日用水量预测是供水系统优化调度的重要前提和关键，其预测精度 直接决定了调度方案是否可行。 本文在系统地总结了国内外城市日用水量预测的研究进展的基础 上，主要完成了以下工作: (1)根据日用水量序列具有较强相关性的特点，引入信息论中可描 述线性和非线性相关的平均互信息法来确定日用水量序列的相关性。 (2)为解决传统日用水量预测模型一般采用主观判断来确定影响因 素的问题，利用改进粗糙集算法对城市日用水量的影响因素进行属性 约简，确定日用水量的主要影响因素。 (3) 以日用水最的主要影响因素和相关日用水量为神经网络的输 入，利用遗传算法优化 BP 网络的权阙值，建立基于粗糙集算法和遗传 BP 神经网络的城市日用水量预测模型。实例分析结果表明，本文提出 的模型具有比基于 BP 网络模型和基于 GA-BP 网络模型更强的预测能 力。 关键词:日用水嚣，粗糙集算沽， BP 神经网络，遗传算法 H H STUDY OF URBAN DAILY WATER CONSUMPTION FORECASTING MODEL BASED ON ROUGH SET AND GENETIC NEURAL NE丁WORK ABSTRACT Along with the rapid expansion of scope and scale of water distribution systems，the complexity of water supply system rises year after year. The traditional operation for urban water supply system face challenge and optimal operation for water supply system is imperative under this si阳ation. Urban dai1y water consumption forecasting is an important prerequisite and the key to optimal management of water supply system，and its accuracy determines whether the optimal operation scheme can be applied to the real water distribution system or not . Research progress of urban daily water consumption forecasting is systematically summarized at home and abroad in this paper，and the main work of this paper is below: (1)According to strong relevance of daily water consumption series， the average mutual infonnation method in information theoηwhich is used to analyze the nonlinear correlation and linear correlation among the daily water consumption demand series. (2)As the influence factors affecting daily water consumption are .IIgenerally determined by subjective judgment ，the improved rough s创 . I I 啦。rithm is used to reduce the factors to determine the principal factors. (3)The principal factors and correlative daily water consumption are usedωthe inputs of neural network，and gen创ic algorithm is introduced to