遗传算法的改进及其在电力系统中的应用研究

发布时间：2018-04-12 13:15

本文选题：电力系统 + 无功优化　；参考：《吉林大学》2014年硕士论文

【摘要】：在电力网络中，实现无功电源的分布，和无功补偿容量的设置，是一个带有大量约束条件的非线性规划问题，对于对大范围供电的供电企业甚至全国范围内的电力能源调配而言，是个非常非常繁杂的过程。遗传算法以初始种群为起点，沿多条路线进行搜索，具有较强的寻优能力，适合非线性离散、多约束、多变量、大规律问题的求解，可以较好的避免“维数灾”的问题，因此在电力系统无功优化领域得到了广泛的应用。但是如果要获得较好的解，那么遗传算法的效率较低，而且也容易产生早熟现象。为此，本文主要针对目前电力系统无功优化的研究现状，针对电力系统无功优化的特点，通过对简单遗传算法的改进来实现对电力系统的无功优化问题进行研究，本文主要的研究内容如下：首先，针对电力系统的无功优化问题，建立以电力系统中，电能损耗最小作为电力系统无功优化问题的目标函数，并且发电机无功越限、节点电压越限作为问题的惩罚函数来进行电力系统无功优化数学模型的研究。其次，针对电力系统无功优化的特点，进行遗传算法的改进，并且对改进遗传算法中的染色体编码算法，选择、变异、交叉等遗传算子，适应度函数的设计以及终止条件的确定等方面，对改进遗传算法的设计进行研究。最后，以一个具体的IEEE 14节点系统为例，利用本文研究的改进的遗传算法，和基本遗传算法，对该电力系统中的无功优化问题进行求解，并且通过两者的对比，从最终的电压控制水平，和降损量两个方面对本文所研究的改进遗传算法的效果进行分析和验证。根据电力系统无功优化的特点，，对简单遗传算法中的编码方式、交叉算子和变异算子的，以及遗传算法的迭代终止条件进行改进，并且在一个具体的IEEE 14节点的无功优化应用中，表明本文所研究的改进遗传算法具有更高的性能和较低的电力系统有功损耗。
[Abstract]:In power network, the distribution of reactive power supply and the setting of reactive power compensation capacity is a nonlinear programming problem with a large number of constraints.It is a very, very complicated process for the distribution of power energy to the power supply enterprises and even the whole country.The genetic algorithm takes the initial population as the starting point and searches along several routes. It has a strong ability to search for optimization. It is suitable for solving nonlinear discrete, multi-constrained, multi-variable and large law problems, which can avoid the problem of "dimension disaster".Therefore, it has been widely used in the field of reactive power optimization in power system.But if we want to get a better solution, the efficiency of genetic algorithm is low, and premature phenomenon is easy to occur.For this reason, this paper mainly aims at the current research situation of reactive power optimization in power system, according to the characteristics of reactive power optimization in power system, through the improvement of simple genetic algorithm to realize the reactive power optimization problem of power system.The main contents of this paper are as follows:Firstly, aiming at the reactive power optimization problem in power system, the minimum power loss is established as the objective function of the reactive power optimization problem in power system, and the reactive power of generator exceeds the limit.As the penalty function of the problem, the node voltage limit is used to study the mathematical model of reactive power optimization in power system.Secondly, according to the characteristics of reactive power optimization in power system, genetic algorithm is improved, and genetic operators, such as chromosome coding algorithm, selection, mutation, crossover and so on, are improved in genetic algorithm.Based on the design of fitness function and the determination of termination conditions, the design of improved genetic algorithm is studied.Finally, taking a specific IEEE 14-bus system as an example, using the improved genetic algorithm and the basic genetic algorithm studied in this paper, the reactive power optimization problem in the power system is solved, and the comparison between the two is given.The effect of the improved genetic algorithm studied in this paper is analyzed and verified from the two aspects of the final voltage control level and the loss reduction.According to the characteristics of reactive power optimization in power system, the coding method, crossover operator, mutation operator and iterative termination condition of genetic algorithm in simple genetic algorithm are improved.And in a specific IEEE 14-bus reactive power optimization application, it is shown that the improved genetic algorithm studied in this paper has higher performance and lower active power loss in power system.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM714.3;TP18

【参考文献】