【作者】 谢维；

【导师】 纪昌明；

【作者基本信息】 华北电力大学 ， 管理科学与工程， 2012， 博士

【摘要】 随着水电能源开发的步伐加快,大规模梯级水电站群的优化调度问题也随之越米越复杂。在梯级水库上下游的水力、电力时空联系,径流不确定性等因素的影响下,传统的系统优化方法与系统分析理论在求解梯级水电站优化调度问题时呈现出一定的局限性。目前,水电站的优化调度规则一般比较简单、笼统,水电站在面临不同来水情况时,这些调度规则的可行性与可操作性较差。因此,完善的梯级水库群优化调度方法与理论研究和可行的优化调度规则制定,是水电调度工作亟待攻克的一项重要课题。基于此,本文以金沙江中游梯级水电站群为研究对象,从理论研究和应用探索两个方面入手,提出了能够有效求解水电站群优化调度问题的粒了群改进算法,针对梯级水电站群联合优化调度问题,探索了一种具体的、可操作性强的水库优化调度规则。本文的研究工作为梯级水电站群的管理、调度等提供了科学的理论基础和可行的优化方法。具体成果如下(1)提出了两种改进粒子群算法。针对基本粒子群优化算法(PSO)在梯级水库群优化调度中的局限性,本文分别从宏观与微观两个层而对粒子群优化算法进行了改进。宏观层面,将文化算法(CA)引入到粒子群优化算法中,提出文化粒子群算法(PSO-CA)。微观层面,将病毒进化机制引入到粒子群优化算法中提出病毒粒子群算法(VPSO)。PSO-CA算法将文化算法(CA)引入到粒子群优化算法中,在PSO算法得到群体空间中各个体的行为的基础上,利用CA算法的信仰空间提炼出群体经验,然后反过来对群体空间陷入局部最优的情况进行监视和改进,提高算法的计算效率。VPSO算法则在计算过程中引入主群体和病毒群体两种群体。主群体在上下代之间纵向传递信息,实现种群的全局寻优；病毒个体通过转录某主个体的基因片段,并反转录给另一个体,在同代个体之间横向传递进化信息,从而克服PSO算法只在群体自身上下代之间传递信息没有在粒子同代之间交换信息的不足。(2)探索了改进粒子群算法在水库优化调度中的应用。不同的水库优化调度问题,对优化方法的要求也不同。因此,梯级水电站群的优化调度需要根据水电系统自身的特征,选择正确的优化方法。本文探索了PSO-CA算法在水电站防洪优化调度与梯级水电站群短期优化调度问题中的应用,以及VPSO算法在梯级水电站群厂间负荷分配与水电站机组负荷分配问题中的应用。通过实例证明,本文提出的两种改进粒子群算法在梯级水电站群优化调度问题中具有明显的优越性。(3)制定了龙盘电站投入前水电站群短期发电优化调度规则。本文根据金沙江中游梯级水电站群三个典型年的资料和大量模拟的径流过程,通过优化算法得到一系列优化调度结果,以此为依据,归纳总结了龙盘电站投入前水电站群短期可行的发电优化调度规则,并给出了具有可操作性的表达形式。(4)制定了龙盘电站投入后梯级水电站群长期联合发电优化调度规则。本文采用了三种不同的调度方式：①常规调度图②蓄能调度图③调度函数来研究梯级水电站群长期联合发电优化调度规则的制定。从不同的侧面对这三种方法进行了分析比较,并进行了较为客观的评价。为水电站制定中长期发电计划与拟定运行指导准则提供了可靠的参考依据。更多还原

【Abstract】 With the accelerated pace of the development of hydroelectric energy, the optimal operation for large-scale cascade hydropower stations is becoming more and more complex and significant. The tranditonal systematic optimizing method and systematic analytical theory may possess some limitations because of the space-time contact of hydraulic power and electric power between upstream and downstream hydropower stations and uncertain factors of the runoff. Nowadays, although many literatures have given the basic rules of reservoir operation, only some general conclusions have been obtained, very few studies have been reported on the clearly different scheduling rules for the different varieties of runoff, and the feasibility of the methods is relatively poor. Therefore, the perfect optimizing scheduling method and theoretical studies for cascade reservoirs and feasible power generation scheduling rules is becoming an important topic. Based on the above, cascade hydropower stations in Jinsha River have been taken as the research object in this paper. Two improved particle swarm optimization algorithm for the optimal operation of the casacade hydropower stations have been proposed based on the two aspects of the theory and application research. Scientific theory foundation and metheod basis for the opearation and management of the hydropower stations have been complemented by the specific and practical scheduling operation rules presented in this paper. Specific results are as follows:(1) Two improved algorithm based on the particle swarm optimization algorithm have been proposed. Particle swarm optimization (PSO) has been modified from the macro level and micro level respectively to overcome the shortage which the PSO possesses when used in the optimizing scheduling of the hydropower stations. In the macro level, particle swarm optimization based on cultural algorithm (PSO-CA) which combines the cultural algorithm (CA) with the particle swarm optimization, and in the micro level, the virus particle swarm optimization algorithm (VPSO) which integrates the virus evolutionary mechanism into the particle swarm algorithm (PSO) is presented in this paper. For PSO-CA, the evolutionary mechanism of particle swarm optimization algorithm (PSO) is guided by cultural algorithm (CA). PSO-CA uses PSO in population space and guides the evolution by shape knowledge and standardization knowledge in belief space. Same examples are used to verify that the PSO-CA algorithm has a better applied prospect for its high reliability and fast operation speed in global optimization. VPSO integrates the virus evolutionary mechanism into the particle swarm algorithm (PSO) and host population and virus population are generated during the evolution. The former transmits genetic information between the different generations which is the same as PSO and the latter carries out the infection operation in the same generation through transcription and reverse transcription. VPSO can effectively inhibit the "premature" phenomenon of particle and accelerate the speed of convergence.(2) The applications of the improved particle swarm algorithm in optimizing operation of reservoir have been researched. Different optimizing scheduling problems of reservoir require the different optimization methods. Therefore, the suitable optimization method of cascade hydropower stations should be chosen according to the characteristics of hydropower system itself. The problems of the optimal flood dispatching of hydropower station and the short-term optimal operation of cascade hydropower stations haven been studied by using PSO-CA, and the problems of the load distribution of cascade hydropower stations and the economic operation of hydropower station have been studied by using VPSO in this paper. The effectiveness of the two improved particle swarm algorithm of cascade hydropower stations in the optimal operation problem also has been verified.(3) Short-term power generation scheduling rules for cascade hydropower stations (before Longpan station startup) of Jinsha River have been formulated. A great quantity of optimal scheduling processes were obtained by calculating the daily runoff process of Jinsha River within three typical years, and a large number of simulated daily runoff processes were obtained using the progressive optimality algorithm (POA) in combination with the genetic algorithm (GA). After analyzing the optimal scheduling processes, the corresponding scheduling rules were given, and the practical formulas were obtained. These rules can make full use of the rolling runoff forecast and carry out the rolling scheduling, and the effectiveness and practical applicability of the rules are testified by a case.(4) Long-term power generation scheduling rules for cascade hydropower stations (after Longpan station startup) of Jinsha River have been studied. The power generations of cascade hydropower stations in Jinsha River were calculated by using three conventional methods of conventional operation chart, energy storage chart and scheduling function, and then were compared with the results obtained by the maximum power generation model and the maximum guarantee output model. By comparing different simulations, the three conventional methods are objectively evaluated herein. The conclusions can provide an important reference for hydropower stations to discover scheduling rules and formulate operation guidelines, and also can provide a reliable basis for making the long-term power generation plan.更多还原