【作者】 王艳；

【导师】 吴学伟；

【作者基本信息】 广州大学 ， 供热、供燃气、通风及空调工程， 2007， 硕士

【摘要】 随着城市的不断发展与扩大，人们面临着前所未有的能源危机和水源危机，因采用人工方法在城市供水系统中调度水资源所带来的问题日益突出，而供水系统优化调度正是有效解决这些问题的措施。本文以广州市自来水公司现有的SCADA调度系统为背景，以该系统提供的用水量、水压、流量等参数的历史数据为研究对象，建立并改进了时间序列模型，得出回归—时序预测模型，建立了广州市宏观模型、一级优化调度模型、二级优化调度模型以及合理的开泵方案。主要研究成果如下：分析了广州市用水量的年、月、日、时变化规律，建立了时间序列模型，并在考虑气候与节假日等外在因素影响的基础上改进时间序列模型，得出回归—时序预测模型。经验证，该模型很好地改善了水量预测的准确性，适合于广州市用水量预测。运用统计分析的方法建立了广州市宏观模型，将管网中若干重要的测压点和水厂流量、压力和几个泵站压力联系起来，建立了它们之间的定量关系。经检验，误差在允许范围之内，为下一步的优化调度提供了必要的约束条件。建立了广州市一级优化调度模型，利用遗传算法解非线性目标规划问题，确定各水厂最优出水量和出厂压力。针对各泵站目标工况点，建立了广州市二级优化调度模型，利用隐枚举法，确定了合理开泵方案。经济比较后发现，优化调度后的方案使总成本降低了7％，具有较强的现实意义。更多还原

【Abstract】 With continuous development and expanse of the city, people are confronted withunprecedented crisis of energy and water resource. Because the problem, which is broughtby scheduling water resource with manual work, is increasingly prominent in water supplysystem, it is optimal scheduling of water supply system that solves these problems.The paper is on the basis of SCADA scheduling system, which is authorized by WaterSupply Company of Guangzhou city. Taking the history data of parameters (including flow,pressure etc) into account, the timing-series model is built and modified, thenregression-timing series model is built; it has also built macroscopical model of Guangzhoucity, primarily optimal scheduling model, secondarily optimal scheduling model and gotreasonable open-close scheme. The primary research fruits as follows:Firstly, water demand changeable regularity of Guangzhou city is analyzed, includingyearly, monthly, daily, hourly water demand, and then, the timing-series model is set up.Considering outside factors influence of weather, vacation and so on, regression-timingseries model is educed. After testing the result, the latter model greatly betters the veracity ofwater demand forecast, so it can be used in Guangzhou city.Secondly, the macroscopical model of Guangzhou city is built with statistical method.Certain important monitoring points in pipe net relate to flow from water stations andpressure from water stations and pump stations, thereby, a quantitative relation is set upamong them. By testing these data, the errors are within permissible range, and the modelcan provide necessary restrictive condition for the following optimal scheduling.Thirdly, the primarily optimal scheduling model is set up, and the nonlinear objectiveprogram is solved with genetic arithmetic. Then the optimal flow and pressure in every waterstations are determined. After that, according to objective status points of each pump stations,the secondary optimal scheduling model is set up, and the reasonable open-close pumpscheme is also determined with enumeration method. It is easy to find that the cost afteroptimal scheduling is lowed 7%, so it has practical meaning.更多还原