| 夏巍,汪石,汪永友.基于改进GRU的城市供水管网流量预测研究[J].安徽建筑大学学报,2023,31(): |
| 基于改进GRU的城市供水管网流量预测研究 |
| Research on flow forecasting methods for urban water supply networks based on improved GRU |
| 投稿时间：2021-11-23 修订日期：2021-12-22 |
| DOI： |
| 中文关键词: 供水管网 流量预测 GRU算法 局部加权线性回归 |
| 英文关键词: water supply network flow data forecast GRU algorithm local weights linear regression |
| 基金项目:巢湖市供水管网优化运行技术及应用示范(2014ZX074055-003-03) |
|
| 摘要点击次数: 1975 |
| 全文下载次数: 596 |
| 中文摘要: |
| 城市供水管网随着城市规模的扩大和人口的增加等因素的变化而越来越复杂，节能降耗的要求更是对供水管网的优化调度提出了新的挑战。供水管网的流量预测，在对管网的优化调度工作中发挥了有效的作用。本文通过对某城市供水管网系统的多个监测节点的流量数据进行分析，以类似于局部加权线性回归的方法来优化GRU算法，构建了基于多监测节点的流量预测模型, 提高了模型的泛化能力。利用某市供水管网监测平台采集的监测节点流量数据进行验证，实验的结果表明基于改进的GRU流量预测模型有着较好的预测精度。 |
| 英文摘要: |
| The urban water supply network is growing increasingly complicated as the city's size and population grow, and the need to preserve energy and cut consumption presents new problems to the optimal scheduling of the water supply network. Flow forecasting of water supply networks is useful in optimizing pipeline network scheduling. In this research, a multi-node flow prediction model is built by analyzing flow data from several nodes in a city water supply network system and optimizing the GRU algorithm using an approach comparable to locally weighted linear regression to increase the model's generalization capacity. Experiment findings reveal that the modified GRU flow prediction model has high prediction accuracy. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
