摘要
文章针对目前大功率静止同步补偿(STATCOM)装置IGBT(Insulated Gate Bipolar Transistor)模块吸收保护电路参数设计中存在的一些问题,对吸收保护电路参数选优问题进行了数学描述,提出了一种采用并行自校正多目标遗传算法来进行大功率STATCOM装置吸收保护电路参数的选型。该算法通过并行进化多种群来确保种群的优良性能和进化的收敛速度,并能够自适应调整遗传交叉概率和变异概率来避免种群的早熟现象,采用阈值制约适应度函数及以不同概率进行的染色体选择操作,可使种群朝多个目标最佳协调点的方向进化,从而得到多目标最优解。应用实例表明采用该方法设计的吸收电路具有较高的工程应用价值。
To solve the problems existing in the design of Insulated Gate Bipolar Transistor (IGBT) snubber circuit parameters for high power STATCOM, the mathematical description for optimal selection of snubber circuit parameters is given and it is put forward that by use of parallel self-correcting multi-objective genetic algorithm the parameters are selected. In this algorithm by means of parallel evolution of multi-population the premium properties of the population and the convergence speed of evolution can be ensured, the precocity of population can be avoided by adaptive adjustment of genetic cross probability and mutation probability. Using threshold limitations of fitness functions and chromosome selection with different probabilities the populations can evolve towards the multi-objective optimal coordination point, thus the multi-objective optimal solution can be obtained. Practical examples show that the snubber circuit designed by this method can better satisfy the requirement from engineering practice.
出处
《电网技术》
EI
CSCD
北大核心
2004年第13期40-44,共5页
Power System Technology
关键词
电力系统
无功功率补偿
大功率静止同步补偿装置
IGBT模块
吸收保护电路
电路参数
优化设计
Distributed parameter networks
Electric insulation coordination
Genetic algorithms
Insulated gate bipolar transistors
Mathematical models
Networks (circuits)
Optimal control systems
Probability