基于鸡群算法的变压器油纸绝缘扩展Debye等效电路参数识别

Parameter Identification for Transformer Oil-paper Insulation Extended Debye Equivalent Circuit Based on Chicken Swarm Optimization Algorithm

为了研究去极化电流和变压器绝缘状态的关系,提出了基于鸡群算法的变压器油纸绝缘扩展Debye等效电路参数识别方法。首先,建立去极化电流的误差函数,计算辨识得到的曲线与实际曲线之间的误差。利用鸡群寻优算法群优化和分类优化的特点,调整电路参数减小与实际曲线误差,解决了粒子群算法易陷入局部最优的缺点。最后通过试验计算表明,应用鸡群寻优算法获得的去极化电流与现场测试得到的数据吻合度高,求得的油纸绝缘扩展Debye等效电路参数能够准确地反映变压器的绝缘状况。

To study the relationship between depolarization current and transformer insulation state,a method was proposed for parameter identification of transformer oil-paper insulation extended Debye Equivalent circuit. Firstly,the error function of the polarization current was established,and the error between the curve obtained through identification and the actual curve was calculated. Then,by using group optimization and classification optimization characteristics of chicken swarm optimization algorithm,circuit parameters were adjusted to reduce the error and overcome the disadvantage that the particle swarm algorithm would likely fall into partial optimization. Finally,experimental calculation showed that the depolarization current obtained through chicken swarm optimization algorithm had a high coincidence with the data obtained from field test,and the parameters obtained for the oil-paper insulation extended Debye equivalent circuit could reflect transformer insulation status accurately.