基于改进飞蛾扑火算法的单时刻参数可变机组组合优化

机组组合属于高维、离散、非凸的混合整数非线性规划问题,具有NPhard特点。提出结合二进制粒子群算法与混沌飞蛾扑火算法的单时刻参数可变机组组合优化方法,将总时刻机组组合问题依次、逐一分解为单时刻启停状态主问题与单时刻经济分配子问题,对主、子问题分别运用二进制粒子群算法与改进飞蛾扑火算法进行交替迭代求解以提升求解速率。运用参数可变策略与优先次序法概率调整策略对算法参数及候选解进行修正,以提升算法运行效率及候选解质量。测试结果表明,本文所提方法具有良好的运算速率及收敛精度,能有效求解大规模机组组合问题。

Optimization of unit commitment with variable parameters in a single hour based on an improved MFO

As a NP hard mathematical program, unit commitment (UC) is a high dimensional,discrete,non convex and mixed integer nonlinear programming. This paper proposes an optimization method for unit commitment with variable parameters in a single hour combined with binary particle swarm optimization (BPSO) and a chaotic moth flame optimization(MFO)algorithm. The method successively resolves the UC problem into the main problem of startup and shutdown status in a single hour and a subproblem of economic allocation in a single hour one by one. Moreover, it uses BPSO and a chaotic MFO algorithm as the alternating iterative method to accelerate the solution. Finally, the paper adopts the parameter variability strategy and probability adjustment strategy based on the priority sequence method to correct algorithm parameters and candidate solutions to improve the operation efficiency of the algorithm and quality of the candidate solution. The test results show that the proposed method,featuring high operation rate and convergence precision, can solve large scale unit commitment.