|
| 摘要: |
| 在深入探讨电力系统机组组合的各种现代智能优化算法的基础上，加以分类总结，详细评述了各种方法所取得的研究成果和存在的不足之处。具体表现在：由于模拟进化算法的随机性，不能保证每次计算都能收敛到全局最优解，同时还存在“早熟”现象；模拟退火算法存在收敛速度慢的缺点；禁忌搜索算法存在对初始解依赖性强和搜索过程只是单对单的操作；人工神经网络的学习训练易陷入局部极值区，同时指出不同的具体问题，网络适合的隐含层数目和节点数目较难确定；模糊优化算法中隶属函数的确定及专家系统中专家的知识、经验和规则的获取都是棘手的问题。 |
| 关键词: 电力系统 机组组合 经济运行 混合整数非线性规则 现代智能优化法 人工神经网络 |
| DOI： |
| 分类号:TM732 |
| 基金项目: |
|
| A survey of modern intelligence optimization methods for unit commitment in electric power systems |
|
YUAN Xiao-hui YUAN Yan-bin ZHANG Yong-chuan
|
| Abstract: |
| The unit commitment plays a more and more important role in the economic operation of electric power system.It is hard to find the global optimization solution owing to the inherent complexity.Based on the study of different optimization techniques,this paper presents the achieve-ments and deficiencies of modern intelligence optimization methods applying to unit commitment in electric power systems .The drawbacks of simulated evolutionary algorithms are that the global optimali-ty can not be always guaranteed because of randomicity and premature con vergence.The convergence speed of simulated annealing is a limiting factor.The drawbacks of Tabu search are its strong sensitivi-ty to initial solution and the only point -to-point search operation.The learning process of artificial neural network is easy to fall into local optimal domain and its numbers of suitable hidden layers and nodes are difficult to be set for a given problem.It is difficult to determine the membership func-tion in fuzzy optimization method and to obtain the knowledge,experience and rules in expert system. |
| Key words: unit commitment,economic operation,optimization methods,mixed integer nonlinear programming |
