基于动态搜索线性混合整数法的机组组合新算法

Linear Mixed Integer Programming Algorithm for Unit Commitment Based on Dynamic Search

针对常用机组组合算法计算量过大的问题,提出了一种基于动态搜索的线性混合整数规划法,该方法在基于分支的线性混合整数规划法的计算框架中引入了动态搜索的概念,采用单独考察与时间相关的约束和功率约束等措施来减少进入分支阶段的数量,在实际运用中,可针对不同规模的系统通过调节计算参数灵活地控制计算时间和计算精度。算例分析结果表明,该方法快速、有效,适合求解大规模电力系统安全约束机组组合问题。

To deal with the large scale unit commitment problems, a new linear mixed integer programming algorithm based on dynamic search technology is proposed. Based on the framework of branch linear mixed integer programming, a new concept, dynamic search, is introduced in this approach. To cut the branches, the constraints concerning time and power are considered separately. For different systems, time of calculations as well as precision of result can be adjusted with appropriate parameters. The result of the test system shows that the presented algorithm is fast, accurate and especially suitable for the unit commitment of large scale power systems.