CAJ | 학술논문

量子粒子群优化算法(QPSO)避免了粒子群算法(PSO)不能保证收敛到全局最优解这个缺点,认为粒子具有量子的行为,并且可以在整个可行解空间进行搜索.无功优化问题是带有离散变量的非线性、不连续、多约束、多变量的复杂优化问题.本文考虑到优化过程中避免陷入局部最优,应用含维变异QPSO算法并结合动态调整罚函数的方法来解决无功优化问题.并对标准IEEE-30节点系统进行仿真计算,并与QPSO、PSO、GA算法进行了比较,表明该算法能够获得更好的全局最优解.
양자입자군우화산법(QPSO)피면료입자군산법(PSO)불능보증수렴도전국최우해저개결점,인위입자구유양자적행위,병차가이재정개가행해공간진행수색.무공우화문제시대유리산변량적비선성、불련속、다약속、다변량적복잡우화문제.본문고필도우화과정중피면함입국부최우,응용함유변이QPSO산법병결합동태조정벌함수적방법래해결무공우화문제.병대표준IEEE-30절점계통진행방진계산,병여QPSO、PSO、GA산법진행료비교,표명해산법능구획득경호적전국최우해.
The question that PSO (Particle Swarm Optimization) can not converge to global optimal solution can be avoided by QPSO (Quantum-behaved Particle Swarm Optimization). QPSO considers that all particles have quantum-behavior and can be searched in the whole area of feasible solution. Ability of global search of QPSO is more excellent than PSO. Reactive power optimization is a complex problem. It is a problem involving discrete variables, nonlinear, discontinuous, multi-constraints and multi-variables. Incorporation with non-stationary assignment penalty function in solving reactive power optimization problem, the proposed QPSODMO (Quantum-behaved Particle Swarm Optimization with Dimension Mutation Operator) method is demonstrated and compared with QPSO ap-proach, PSO approach and GA approach on the standard IEEE30-bus system. The investigations reveal that the proposed method is efficient in solving reactive power optimization problem. And by IEEE30 testing, reactive power optimization problem solved by QPSODMO can find better solution and have less loss than QPSO and the other PSO approach. Simulating algorithm shows feasibility and validity of the solution.