含双馈型风电场的电力系统无功优化策略研究

发布时间：2018-04-03 22:17

本文选题：双馈风力发电场　切入点：电力系统无功优化　出处：《兰州理工大学》2017年硕士论文

【摘要】：能源是一个国家发展的基本推动力,面对当今社会化石能源日渐短缺,空气质量逐年下降,雾霾问题日益严重的现状,各国都将发展重点放在了新能源上。并网发电的新能源主要包括太阳能、风能、水能、核能、生物能源等。其中风能因其存在范围广且可开发利用程度高,成为新能源发展的主要发力点。截止2010年年底,我国风电累计装机容量已经跃居为世界第一,风力发电展现出良好的发展前景与应用前景。伴随着风力发电研究发展的不断深入,风力发电将为国民经济发展提供更多优质的电能。我国风力发电场采用较多的是双馈异步风力发电机,因此论文选取其作为研究对象,重点分析了双馈异步风力发电机的基本类型和出力特征,并用数学方法分析建立了双馈异步风力发电机的稳态数学模型以及动态数学模型。在此基础上,论文阐述了双馈异步风力发电机的有功功率,电场电压以及其无功功率之间的数学关系,并建立了含双馈异步风电场的电力系统的潮流计算模型。其次,论文综合了粒子群优化算法(Particle Swarm Optimization,PSO)以及和声搜索算法(Harmony Search Algorithm,HSA)的优点,即将和声搜索算法能够有效跳出局部最优的特性与粒子群算法收敛具有方向的特性结合。将和声库中的元素视作粒子群中的粒子,每次迭代时,首先利用粒子群算法对和声记忆库中的记忆元素进行一次寻优,然后将寻优后的和声记忆库元素带入改进和声搜索算法。论文选取了四个常用的多目标收敛测试函数对所提算法的寻优性能进行了检测,并与常见的智能算法进行了收敛对比测试,测试结果说明了论文所改进的算法具有较好的收敛精度以及收敛速度。最后,论文针对含双馈型风力发电机的风电场接入电力系统时的无功补偿问题,建立了有功网损最小以及节点电压偏差最小的双目标无功优化模型,采用上文改进的粒子群优化和声搜索算法,对含风电场的电力系统进行无功优化研究。将论文提出的改进粒子群优化和声搜索算法应用于含风电场的IEEE30节点电力系统进行无功优化,并与粒子群算法以及非支配排序遗传算法(NSGA-Ⅱ)算法的无功优化结果进行比较。实验结果表明,论文所改进的粒子群优化和声搜索算法具有较好的收敛精度以及较快的收敛速度,在使用该算法对电力系统进行无功优化之后,系统内各节点电压幅值得到了不同程度的改善,降低了有功网损值,提高了含风电场电力系统的电能质量以及电力系统对风电的接纳能力,有利于促进风力发电的进一步发展。
[Abstract]:Energy is the basic driving force of the development of a country, in the face of the growing shortage of fossil energy in today's society, declining air quality, present situation of increasingly serious haze problems, all countries will focus on the development of new energy power generation. New energy including solar energy, wind energy, hydropower, nuclear energy, bio energy, wind energy. Because of the wide range and extent of development and utilization, has become the main force of the new energy development. By the end of 2010, China's total wind power installed capacity has been ranked as the world's first wind power, show the development prospects and good application prospects. With the development of wind power research deepening, wind power will to provide more high-quality electricity for the national economic development. China's wind power field is used more doubly fed asynchronous wind generator, so this paper chooses it as the research object, analyzes the The basic types of doubly fed asynchronous wind generator and output characteristics, and the establishment of the mathematical model of doubly fed asynchronous wind generator and dynamic mathematical model by mathematical analysis method. On this basis, this paper expounds the active power of DFIG, the mathematical relation between electric voltage and reactive power, and the establishment of the power system with doubly fed induction wind power flow calculation model. Secondly, the particle swarm optimization algorithm (Particle Swarm Optimization, PSO) and harmony search algorithm (Harmony Search Algorithm, HSA) advantages, will search characteristics and acoustic characteristics and particle swarm optimization algorithm can effectively jump out of local optimum with direction the combination of the elements in the library. And as the particles, each iteration, using particle swarm optimization algorithm for the harmony memory memory element In a search, the harmony memory elements will then be optimized into improved harmony search algorithm. This paper selects four common multi-objective function convergence test to test the performance of optimization algorithm, and intelligent algorithms and the common convergence comparison test, the test results show the improved algorithm has better convergence accuracy and convergence speed. Finally, reactive power compensation of wind farm power system according to the content of double fed wind generator is established, the minimum network loss and node voltage deviation of minimum double target reactive power optimization model, using the improved particle swarm optimization and above search algorithm of power system containing wind farms for reactive power optimization. The proposed improved particle swarm optimization and search algorithm is applied to wind power system IEEE30 node Reactive power optimization, and the particle swarm algorithm and non dominated sorting genetic algorithm (NSGA-) algorithm for reactive power optimization results were compared. The experimental results show that the improved particle swarm optimization and search algorithm has better convergence precision and fast convergence speed, after the use of the algorithm in power system without reactive power optimization, the system of the node voltage has been improved to some extent, reduce the power loss, improve the ability to accept the power system with wind electric power quality and power system of wind power, to promote the further development of wind power generation.

【学位授予单位】：兰州理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM614;TM714.3

【参考文献】