摘要
针对粒子群算法在求解油纸绝缘等效电路模型参数过程中容易陷入局部最优的问题,结合混沌理论提出一种油纸绝缘极化等效电路参数计算方法。根据混沌算法具有遍历性和不重复性的特点,在粒子群算法中加入混沌扰量,使得算法不陷入局部最优解,从而辨识油纸绝缘等效电路参数。计算结果表明,应用混沌粒子群改进算法获得的回复电压极化谱能够很好地与现场测试得到的回复电压极化谱相吻合。通过所提算法求解得到的油纸绝缘等效电路模型能够准确地反映油纸变压器的绝缘状况,为后续诊断油纸绝缘的老化情况奠定重要基础。
Aiming at the problem of the particle swarm algorithm easily into the local optimum in the process of solving oil-paper insulation equivalent circuit model parameters,a new method of calculating the oil-paper insulation equivalent circuit parameters was proposed combined with the chaos theory.According to the chaos algorithm characteristics of ergodicity and no repetitiveness,combining the chaos disturbance with the particle swam algorithm,which does not fall into local optimal solution,oil-paper insulation equivalent circuit parameters was identified.The calculation results show that the return voltage polarization spectrum obtained by improved particle swarm optimization algorithm can be matched better with the return voltage polarization spectrum obtained by field test.The oil-paper insulation equivalent circuit model based on the new algorithm can reflect the insulation condition of oil-paper insulation transformer accurately,which laid an important foundation for the subsequent diagnosising oil-paper insulation aging.
出处
《电机与控制学报》
EI
CSCD
北大核心
2014年第8期62-66,共5页
Electric Machines and Control
基金
国家自然科学基金(61174117)
关键词
等效电路
参数辨识
回复电压
混沌粒子群
极化谱
equivalent circuit
parameter identification
return voltage
chaos particle swarm
polarization spectrum