【作者】 何中杰；

【导师】 汪雄海；

【作者基本信息】 浙江大学 ， 电气工程， 2009， 博士

【摘要】 城市排水系统属多输入、多目标的广地域分布复杂非线性耦合系统,不合理的排放控制方式易造成泵站高能耗及污水溢出污染。针对城市分布式泵站排水系统综合节能与协调优化控制问题,本论文主要研究内容和成果如下：(1)建立了基于机泵群控制、具有时滞和相互链级关系的分布式泵站排水控制系统模型。依据机泵相似律推导出调速特性模型通式,以及基于指数律的机泵调速特性数值模型。探讨了一种基于机泵调速特性模型的实时流量优化测算方法,通过实验检验了该方法的有效性。(2)以分布式泵站的节能运行为目标,分析了降低静扬程和动扬程的综合节能机理,并给出了适用于不同工况的两种机泵调速节能优化方法：1)针对扬程变化幅度大且不易控制、或扬程与机泵效率比值不满足递减规律的工况,推导了机泵保持等效运行的充要条件,从而得到一种跟随静扬程变化的自适应高效控制法；2)对于静扬程易自主调节且扬程效率比可满足递减规律的一类工况,综合考虑机泵功率与效率,可得依据扬程效率比的节能法。通过实验比较了两种节能控制法的节能效果。(3)针对广地域城市排水系统的协调优化需求,以及引入网络控制后带来的控制品质下降问题,构造了城市排水系统的分层协调控制结构。推导求证了区域泵站控制系统可控和可观测的充要条件。给出了一种初值匹配参数自适应辨识方法,以及结合该方法的多模型预测函数控制(PFC)法；仿真结果表明,将该方法用于区域泵站的闭环控制,系统辨识效果优于采用ARMAX模型的参数辨识方法,且具有较好的控制鲁棒性。(4)为改善广地域分布系统工况差异性带来的合流制溢流/分流制溢流(CSO/SSO)问题,提出了一种适用于分布式泵站排水系统的流量协调优化调度方法。引入相互作用预测法,结合流量协调优化调度法,将最优控制转化为最优控制和协调优化调度的多目标综合协调优化。仿真结果表明,当入流量不超过各级泵站的排水能力之和时,综合协调优化能根据工况差异有效利用各分布泵站排水能力,确保全局无溢流,不受时滞和决策变量变化的影响。(5)设计了一个基于开放标准以太网和Internet的多回路并行C/S网络结构,实现了城市分布式泵站排水网络控制系统中多类型、多回路控制器的网络协调优化调度,为基于分层控制结构的分布式泵站排水系统综合协调优化控制提供了软件平台。更多还原

【Abstract】 Urban drainage systems are non-linear coupled system with multi-input and multi-objective complexity of the wide geographical distribution. Unreasonable emission control method will cause high energy consumption of pumps and pollution of sewage overflow. Aiming at the problem of comprehensive energy saving and coordinative optimization control of distributed pumping stations in urban drainage systems, the main research topics and contributions of the dissertation are:(1) A pump group control based urban drainage control systems model was established to describe the urban drainage systems containing unexpected pipeline delay and mutual relations between pumping stations. According to the affinity laws of pump, a general formula for variable-speed characteristic of pump group was derived, then an exponential law based numerical model of variable-speed pump group was obtained. Based on the proposed numerical model, an optimal real-time flow measurement method was discussed. The effectiveness of the measurement method was tested through experiments.(2) Take the energy-saving operation of distributed pumping stations as the target, the energy-saving mechanism that integrates reducing static head and dynamic head was analyzed. Then, two optimize energy-saving methods for variable-speed pump which are applicable to different conditions were presented: 1) A high efficient self-adaptive control method follows the variation of static head. This method is suitable for the situation that large range of static head is difficult to control and the ratio of head and efficiency is not satisfied with the law of diminishing.2) An integrated optimal control method considering the power and efficiency of pump simultaneously takes the ratio of head and efficiency as the performance indicator. This method is suitable for the situation that static head is easy to self-regulation and the ratio of head and efficiency is satisfied with the law of diminishing. The energy-saving effect of two method was compared through experimentation.(3) Addressing the optimize coordination demand of large geographically distributed urban drainage systems, and the decline in the quality control after the introduction of networked control, A hierarchical coordinate urban drainage networked control systems was presented. An adaptive parameter estimation law with initial value matching was proposed. Based on the proposed law, a multi-model switching control method integrated of predictive functional control (PFC) was developed. The simulation results show that it has better performance of system identification that applying the proposed parameter estimation law to Regional Station closed-loop control systems instead of using ARMAX model, and the proposed control method has good robustness.(4) In order to improve Combined Sewer Overflow / Sanitary Sewer Overflow (CSO / SSO) problem brought about by differences in operating conditions of spatially distributed drainage systems, a coordinated discharge scheduling method was discussed,which is adapt to distributed pumping stations in urban drainage systems. By the discharge scheduling method and introduction of the interaction prediction method, optimal control will be converted into multi-objective optimization integrating optimal control and comprehensive scheduling. The simulation results show that coordinative optimization is able to effective use of the capacity of the distributed drainage pumping stations in terms of conditions diversity when the inflow does not exceed the total capacity of drainage pumping station in every level, which ensures overall none sewer overflow. This coordinative optimization method does not affected by the changes of pipeline delay and decision-making variables.(5) A multi-channel, parallel network architecture based on the open standards Ethernet and Internet was designed. In the basis of the architecture, coordinative network scheduling of multi-type and multi-loop controllers of distributed pumping stations in urban drainage networked control systems was implemented, which provide a software platform for optimal hierarchical control of the coordination of distributed pumping stations in urban drainage systems.更多还原