配水管网测点压力数据的建模与预报

Model-building and Forecasting to Pressure of Monitoring Node of Water Distribution Network

给水管网测点压力为一串时序样本数据,本文引进时序建模方法对样本数据进行相关分析,特别结合实例对四种建模方案作了分析比较,结果表明带受控项同时反映周期特性的模型精度最高,完全满足工程要求,本文还对测点压力在线建模及其预报算法作了探讨,取得了满意的结果,作者期望对配水管阿工况作出有效估计,为配水系统的优化调度与节能创造条件。

The monitoring node pressure of water distribution network is a time series sampling data. This paper introduces time series model-building method to the sampling data for correlation analysis, especially combining realistic example, we devolop four model-building schemes for analysis and comparision, the conclusion is that the model with control terms and periodic character has high precision,can fully satisfy the engineering demand. This paper approaches on-line model-building and forecast about monitoring node pressure,obtainning satisfied results. We expect to cany out available estimation to operation status of water distribution system and provide convenience for optimal control to water supply system.