求解机组组合问题的领域搜索法  

Unit Commitment via Local Search Point Method

作　　者：黎静华[1] 韦化[1] 

机构地区：[1]广西大学电气工程学院,广西壮族自治区南宁市530004

出　　处：《中国电机工程学报》2008年第13期33-40,共8页Proceedings of the CSEE

基　　金：国家自然科学基金项目(59867001);教育部高等学校博士学科点专项基金项目(20060593002)~~

摘　　要：机组组合问题是电力系统优化运行的一个难点,理论上难以得到其最优解。该文提出用邻域搜索(local search,LS)和内点(interior point,IP)法相结合的算法(LS-IP)解决机组组合(unit commitment,UC)非确定多项式时间(nondeterministic polynomial,NP)难问题。定义邻域的结构,并提出一种邻域的调整方法,可处理各项约束条件,保证结果的可行性。用非常小的解邻域空间代替原来庞大复杂甚至难以求解的离散空间。充分利用内点法收敛性好、精度高的优势,提高其计算速度。对100台机组24时段仿真结果表明,CPU计算时间仅为原来的4s,所耗费用却大大降低;同时该方法收敛速度快、精度高,尤其适合于求解大规模机组的组合问题。Unit commitment is an important and difficulty aspect of optimal operation of power systems. Because it belongs to be mixed variables programming that is difficult to find the optimal solution in mathematics. This paper presented a new method combining of neighbor search and interior point to solve unit commitment （UC） nondeterministic polynomial （NP）-hard problem. The structure of neighborhood is defined and a new way to adjust neighborhood was put forward in which various restrictions can be dealt with to ensure the result feasibility. The large and complex discrete space which is even difficult to obtain solutions will be replaced by a very small solution neighborhood. The advantages of interior point method, for example with good astringency and high precision, will be fully played to accelerate the speed of calculation. 100 units and 24 hours systems were tested to verify the presented method. The CPU time cost is only 4s while the financial cost is very low, which have shown the proposed method is very promising for large scale application.

关 键 词：混合整数规划 机组组合 内点法 优化运行 电力系统 

分 类 号：TM76[电气工程—电力系统及自动化]