随机需求下的梯级泵站日优化调度

Optimal Daily Scheduling of Multi-stage Pumping Stations Based on Stochastic Demand

梯级泵站的优化运行能极大地提高系统的经济效益.根据日调度预见期较短的特点,采用核函数回归预测次日的系统需水量,带宽的选择采用一种简单的基于经验的规则确定.基于预测日供水量,提出了一种考虑区段流量延迟时间的影响和峰谷电价差异的优化调度模型.对东深供水工程进行了优化计算,结果表明所建立的模型能减少系统的抽水成本.

The optimal operation of multi-stage pumping stations can greatly increase the economical ef- ficiency of water supply system. Due to a short lead time in daily operation, a kernel regress was used to forecast the day-ahead water demand of system. The bandwidth of kernel function was simply cho- sen by a rule of thumb. Based on the forecast value, an optimal model taking the influence of time de- lay of discharge between the pumping station and difference of time depend electricity price into ae- count was proposed. A simulation was performed for Dong-Shen water supply system, and test results shows that the proposed model can decrease the pumping cost of system.