CAJ | 학술논문

粒子群优化(PSO)算法中的粒子根据全局最优粒子来移动自身位置进行搜索,但当某一粒子连续多次被选为全局最优粒子的时候,整个群体的粒子就会快速收敛于该最优粒子,陷入局部最优.为此,提出了变异动态粒子群优化(MDPSO)算法.采用惯性权重变异的思想,当某粒子连续被选为全局最优粒子时,就使一部分粒子的惯性权重以指数速度增长,使粒子跳出局部最小,继续全局寻优.并把改进的粒子群优化算法和BP神经网络相结合,应用于变压器故障检测中.实验表明,与常用的粒子群优化算法相比,用改进的粒子群优化算法优化BP神经网络具有更好的性能,在变压器故障检测中能够获得更高的检测精度.
입자군우화(PSO)산법중적입자근거전국최우입자래이동자신위치진행수색,단당모일입자련속다차피선위전국최우입자적시후,정개군체적입자취회쾌속수렴우해최우입자,함입국부최우.위차,제출료변이동태입자군우화(MDPSO)산법.채용관성권중변이적사상,당모입자련속피선위전국최우입자시,취사일부분입자적관성권중이지수속도증장,사입자도출국부최소,계속전국심우.병파개진적입자군우화산법화BP신경망락상결합,응용우변압기고장검측중.실험표명,여상용적입자군우화산법상비,용개진적입자군우화산법우화BP신경망락구유경호적성능,재변압기고장검측중능구획득경고적검측정도.
Particle Swarm Optimization (PSO) algorithm searches the best solution by making particles moving around the search space according to the global best particle. But when one particle is selected as the global best particle continuously, the other particles will converge at the global best particle repeatedly, which makes the particle swarm fall into local optimization. The authors presented Mutational Dynamic Particle Swarm Optimization (MDPSO) algorithm. A part of particles' inertia weight would mutate when one particle was selected as the global best particle continuously, which could make the part of particles jumping out of the local optimization and keeping searching in the whole solution space. Otherwise, the authors combined MDPSO and BP neural network and applied it to the diagnosis of power transformer. The experimental results show that the proposed approach has a better ability in optimizing BP neural network and in terms of diagnosis accuracy.