基于改进粒子群算法的含分布式电源的配电网优化重构研究

发布时间：2018-09-08 20:01

【摘要】：配电网是国民经济和社会发展的重要公共基础设施。配电网重构是确保电力系统安全、稳定、可靠运行的有效方法之一。本文旨在降低配电网网损、提高系统经济性,采用一种改进的动态重构数学模型,即在固定时间段内考虑系统的多种负荷运行方式,首先单独计算各负荷方式下的系统网损,然后以各负荷方式运行时间占总时间的比例为权重,来考虑系统在多种负荷方式下的综合网损进行重构。采用这种动态重构模型,既能够较大程度的反映固定时间段内负荷和网损的变化情况,又能避免采用在实际操作上具有很大局限性的分时段优化策略,使得模型在理论上更贴合实际,同时这种模型也继承了静态重构模型计算简便、高效的特点。本文介绍了传统PSO算法和3种惯性权重的改进方法,通过概率分布图详细地比较了均匀变异、高斯变异、多项式变异3种变异算子在全局搜索和局部搜索性能上的优劣。针对传统PSO算法往往有容易陷入局部最优值的缺陷,本文将均匀变异算子引入二进制PSO算法中,有效地解决了算法早熟的问题。同时为了减少因均匀变异算子在算法后期可能降低寻优精度所带来的不利影响,采用线性递减的惯性权重来确保算法后期的寻优精度,此外还采用了效率更高的编码策略使得运算过程中不可行解的产生率大大降低。最后,通过标准IEEE 33节点算例,将改进二进制PSO算法与传统PSO算法的性能作比较,以此验证改进二进制PSO算法的实用性,之后分别通过加入DG后的IEEE 33节点算例和IEEE 16节点算例,验证改进二进制PSO算法对于加入DG的配电网同样具有良好的运算性能。
[Abstract]:Distribution network is an important public infrastructure for national economy and social development. Distribution network reconfiguration is one of the effective methods to ensure the safe, stable and reliable operation of power system. The purpose of this paper is to reduce the distribution network loss and improve the system economy. An improved dynamic reconfiguration mathematical model is adopted, that is, considering the various load operation modes of the system in a fixed time period, the system network losses under each load mode are calculated separately at first. Then taking the proportion of operation time of each load mode to the total time as the weight, the comprehensive network loss of the system under various load modes is considered to be reconstructed. This dynamic reconstruction model can not only reflect the change of load and network loss in a fixed time period to a large extent, but also avoid the use of time-divided optimization strategy, which has great limitations in practical operation. It makes the model more practical in theory and inherits the simple and efficient calculation of static reconstruction model. This paper introduces the traditional PSO algorithm and three improved methods of inertial weight, and compares the global search and local search performance of three mutation operators: uniform mutation, Gao Si mutation and polynomial mutation in detail by probability distribution map. In view of the defect that the traditional PSO algorithm is easy to fall into the local optimal value, this paper introduces the uniform mutation operator into the binary PSO algorithm, which effectively solves the premature problem of the algorithm. At the same time, in order to reduce the adverse effect of uniform mutation operator in the later stage of the algorithm, the linear decreasing inertia weight is used to ensure the optimization accuracy in the later stage of the algorithm. In addition, a more efficient coding strategy is used to reduce the generation rate of infeasible solutions. Finally, the performance of the improved binary PSO algorithm is compared with that of the traditional PSO algorithm through an example of standard IEEE 33 node, which verifies the practicability of the improved binary PSO algorithm. After that, the improved binary PSO algorithm is proved to have good performance for distribution network with DG by the IEEE 33 node example and the IEEE 16 bus example after the addition of DG.
【学位授予单位】：广西大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TM73;TP18

【相似文献】