长距离输水多级泵站系统运行调度优化建模与决策方法研究

Long distance multistage pumping station system consumes a large amount of energy at runtime. If scheduling optimization operation is achieved, the operation cost of pumping station system could be reduced obviously, but so far it has not been achieved because of the complexity..Multistage pumping station system includes main pump units, auxiliary equipments, power transmission and transformation equipments and water facilities. The relationship is complicated among each part, which affects energy consumption of the system. For multistage pumping station system, each part of the energy losses would be researched. Aiming at different water supply demands, optimal models with many factors would be reasonably established by Analytic Hierarchy Process, and the optimizing objects would be more comprehensive. For Hybrid Particle Swarm Optimization algorithm, the reliability, convergence and stability would be analyzed, also the algorithm improvements would be researched. The decision methods would be determined based on the models. Multi-layer iterative method would be proposed to solve the models, also waterlevel - flow database of interstage water transfferring and optimization scheme database inside pumping stations would be established to reduce calculation amount and improve calculation accuracy. A method of overall coordination optimization would be discussed to achieve simultaneous optimization for each stage, but there are certain difficulties. The optimal operation schemes would be calculated for a typical long distance multistage pumping station system. .According to the characteristics and requirements of multistage pumping station, the project tries to establish models and put forward decision methods from the view of system. The calculating speed and the precision could be higher. The results could enrich the optimization theories and achieve scheduling optimization operation for long distance multistage pumping station system.

长距离输水多级泵站系统运行耗能巨大，实现优化调度，可显著降低运行成本，但因其复杂，至今未能实现。.多级泵站系统包括主机组、辅助设备、输变电设施和输水设施。各部分关系复杂，影响系统能耗。项目研究长距离输水多级泵站系统各部分能量损失，针对不同供水要求，应用层次分析法合理构建考虑多因素的多级泵站系统运行优化模型，优化对象更加全面。研究混合粒子群算法的可靠性、收敛性、稳定性，并改进算法。针对构建的模型，确定决策方法。提出多层次迭代优化方法求解模型，建立级间输水水位—流量数据库与站内运行优化方案数据库，减小计算量，提高计算精度；探讨建立整体协调优化方法求解模型，实现系统各个梯级同时优化，但有一定难度。计算典型长距离输水多级泵站系统优化运行方案。.项目针对多级泵站特点和要求，从系统角度构建运行调度优化模型，提出决策方法，计算速度快、精度高。成果能够丰富优化理论，实现长距离输水多级泵站系统运行优化调度。

项目针对长距离输水多级泵站系统，研究了泵站输变电、主电动机、传动装置、主水泵装置和辅助设备等各部分运行能耗，分析了能耗影响因素，确定了系统能耗定量计算方法。. 针对不同供水要求，以系统运行能耗最低为优化目标，应用层次分析法分别建立了单级泵站系统运行优化调度模型、单线输水多级泵站系统运行优化调度模型和多线输水多级泵站系统运行优化调度模型。. 对粒子群算法参数进行了调试与选择，保证算法的可靠性，采用标准偏差进行评价。采用开口向下抛物线递减的方式更新惯性权重值，引入模拟退火算法，增强混合算法的全局搜索能力。研究了算法对其控制参数的敏感程度。采用不可行规则进行处理约束条件。. 采用两种方法求解优化调度模型。多层次迭代优化方法将模型分解成若干个层次，从最内一层开始计算，比较判断每层最优结果，并将其逐级向上层迭代，直至完成最外一层次的计算，最后综合确定出系统运行调度优化方案。基于混合PSO算法的整体协调优化方法将输水线路流量、泵站扬程、单机流量等物理量直接定义为变量，代入算法流程中进行计算，实现多级泵站系统中各个梯级同时优化，确定系统运行优化方案。. 以南水北调东线泗阳站泵站系统为对象，恒定流量变角优化运行方案、加大流量变角优化运行方案分别较设计角度运行方案节能0.14%-3.18%、14.62%-30.87%。以南水北调东线洪泽湖至骆马湖段多级泵站系统为对象，对于徐洪河线单线输水多级泵站系统，洪泽湖、骆马湖水位分别为12.5 m、23.5 m时，级间扬程与站内机组优化运行方案，较不优化方案节能3.77%-12.74%。整体协调优化方法计算精度更高，较多层次迭代优化方法节约系统能耗1.7%左右。对于徐洪河线、中运河线双线输水多级泵站系统，洪泽湖、骆马湖水位分别为13 m、23 m时，两线流量优化分配、级间扬程优化分配、站内机组变角优化运行方案较不优化方案节能8.68%-13.96%。. 成果丰富了多级泵站系统运行优化理论，实现了长距离输水多级泵站系统运行优化调度。