考虑风电场并网的电力系统无功优化

发布时间：2018-01-21 15:16

本文关键词： 系统无功优化风电并网混沌粒子群算法蛙跳算法　出处：《山东大学》2015年硕士论文　论文类型：学位论文

【摘要】：我国风力资源丰富,作为绿色、无污染、可再生的清洁资源之一,风电的发展给电力行业带来了无限前景。由于风力的特点,风电的输出具有随机性和波动性,在大规模的风电并网时会给系统的稳定性带来冲击。对于电力系统来说,无功功率的平衡影响着系统的电压质量,是决定系统能否保持稳定运行的重要条件。电力系统的无功优化是一个极具复杂性的问题,多变量、多约束。大规模风电接入电网后,给无功优化带来了新的挑战,风速难以预测,输出功率数据难以得到；传统计算潮流的算法对于风电节点不再适合；用于优化问题的各种算法都存在一定的缺点,亟需改进。本文针对风速难以预测的问题,建立了满足精度要求、计算方便的风电场功率输出模型。以Weibull分布近似表征风速,利用风电场景概率模型,计算得到风场输出功率。选用牛顿-拉夫逊法进行系统潮流计算,考虑风电并网节点的特殊性,该算法对于并网节点只需修改风电节点相应的雅克比矩阵元素。无功优化的过程就是一个最优潮流求解的过程,选择合适的算法对于最优解的获取至关重要。本文选用获得广泛认可的粒子群算法作为寻优算法,以系统的有功网损最小作为目标函数,同时加入对节点电压越限和发电机无功越限的惩罚项。针对粒子群算法易陷入局部最优的缺点,引入混沌粒子群算法。在混沌粒子群算法的基础上,本文提出了一种新的算法——蛙跳混沌粒子群算法(Frog-Chaotic Particle Swarm Optimization, F-CPSO)。蛙跳算法作为一种新的启发式进化算法,具有良好的计算性能和全局寻优能力。为克服粒子群算法易陷入局部最优的缺点,F-CPSO在混沌粒子群算法的基础上做了改进：对初始种群以粒子适应度函数值的大小为准则将种群分组,在族群和全局内同步寻优；在速度更新公式中加入对族群最优粒子的学习因子；每一次的迭代过程中,随机产生一个新的粒子替换适应值最差的粒子,增加种群的多样性。为验证改进算法的有效性,本文以IEEE-30节点系统为例进行了仿真,在IEEE-30节点系统加入风电节点。采用MATLAB语言进行编程仿真,并与标准粒子群算法(SPSO)进行对比,经过仿真的结果分析验证,改进后的粒子群算法能够有效地降低系统的有功网损,提高系统的电压合格率,相比于标准粒子群算法,改进后的粒子群算法具有更好的全局收敛性能。
[Abstract]:As one of the green, pollution-free and renewable clean resources, the development of wind power in China brings infinite prospects to the power industry. Because of the characteristics of wind power, the output of wind power has randomness and volatility. The stability of the power system will be impacted by large-scale wind power grid connection. For the power system, the balance of reactive power affects the voltage quality of the system. Reactive power optimization of power system is a very complex problem, multivariable, multi-constraint, large-scale wind power connected to the power network. It brings new challenges to reactive power optimization, wind speed is difficult to predict, output power data is difficult to obtain; The traditional algorithm for calculating power flow is no longer suitable for wind power nodes. All kinds of algorithms used in optimization problems have some shortcomings and need to be improved. In order to solve the problem that wind speed is difficult to predict, this paper establishes a method to meet the precision requirements. The wind speed is approximately represented by Weibull distribution, and the probability model of wind power scene is used. The output power of wind field is calculated. Newton-Raphson method is used to calculate the power flow of the system, and the particularity of the connected node of wind power is considered. The algorithm only needs to modify the corresponding Jacobian matrix elements for grid-connected nodes. The process of reactive power optimization is a process of solving the optimal power flow. It is very important to select the appropriate algorithm to obtain the optimal solution. In this paper, the widely accepted particle swarm optimization algorithm is chosen as the optimization algorithm, and the minimum active power network loss of the system is taken as the objective function. At the same time, the penalty items of the node voltage overrun and generator reactive power are added. Aiming at the disadvantage of particle swarm optimization, chaotic particle swarm optimization algorithm is introduced. Based on chaotic particle swarm optimization algorithm, chaotic particle swarm optimization algorithm is introduced. In this paper, a new algorithm, Frog-chaotic Particle Swarm Optimization, is proposed. As a new heuristic evolutionary algorithm, F-CPSOL has good computational performance and global optimization ability. In order to overcome the shortcomings of particle swarm optimization (PSO), it is easy to fall into local optimization. F-CPSO is improved on the basis of chaotic particle swarm optimization algorithm: the initial population is grouped according to the size of the particle fitness function, and the population is optimized synchronously within the population and the whole world; Adding the learning factor to the population optimal particle in the velocity update formula; In each iteration, a new particle is randomly generated to replace the worst particle, increasing the diversity of the population. In this paper, IEEE-30 node system is taken as an example, wind power node is added to IEEE-30 node system, and MATLAB language is used to program simulation. Compared with the standard particle swarm optimization (SPSO), the improved particle swarm optimization algorithm can effectively reduce the active power loss of the system and increase the qualified rate of the system through the analysis of simulation results. Compared with the standard particle swarm optimization algorithm, the improved particle swarm optimization algorithm has better global convergence performance.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TM614;TM714.3

【参考文献】