摘要
针对大规模电力系统机组组合问题,提出了基于集结投影次梯度方法的分解协调算法。首先在上层通过拉格朗日松弛方法将原问题分解为多个子问题,从而减小了求解问题的复杂度,避免了维数灾问题,同时显著降低了计算时间,使得原问题可以在多项式时间内求解,随后下层子问题采用动态规划方法很容易求最优解。算例仿真结果表明,所采用的集结投影次梯度方法调整拉格朗日乘子,避免了传统次梯度方法振荡现象严重的缺点,同时加快了收敛速度,得到了令人满意的机组组合方案。
Unit commitment of large power system is very difficult to solve because of its non - convex, discrete, and NP - hard combinatorial characteristics. This paper proposes an improved Lagrangian relaxation algorithm based on aggregative projection subgradient method,which decreases the dimensions of the primal problem by decomposing the large problem into multi - sub problems, then decreases the computational time and attains the sub - optimal solu- tions in polynomial - time. The simulation results show that using aggregative projection subgradient method to adjust Lagrangian multipliers can reduce the oscillation of subproblem solutions in the primal space, improve the conver- gence of the dual problem and find a satisfactory scheme of unit commitment.
出处
《计算机仿真》
CSCD
2008年第2期245-247,302,共4页
Computer Simulation
基金
国家863科研项目发展基金资助项目(2002AA517020)
关键词
机组组合
拉格朗日松弛
集结投影次梯度
Unit commitment
Lagrangian relaxation
Aggregative projection subgradient
