中国电力 ›› 2021, Vol. 54 ›› Issue (7): 109-116.DOI: 10.11930/j.issn.1004-9649.201908012

• 电网 • 上一篇    下一篇

一种用于电流互感器铁心J-A磁滞参数识别的改进方法

周仕豪, 潘本仁, 万勇, 王冠南, 张妍, 桂小智   

  1. 国网江西省电力有限公司电力科学研究院,江西 南昌 330096
  • 收稿日期:2019-08-01 修回日期:2020-01-04 发布日期:2021-07-12
  • 作者简介:苟竞(1988-),男,博士,高级工程师,从事电力系统规划和新能源并网设计工作,E-mail:scu_goujing@163.com;况理(1996-),男,硕士研究生,从事低惯量电力系统优化调度与运行研究,E-mail:likuang12138@qq.com;文云峰(1986-),男,通信作者,博士,副教授,从事低惯性电力系统规划、运行与控制研究,E-mail:yunfeng.8681@163.com
  • 基金资助:
    国家电网公司科技项目(521820180017)

An Improved Method for J-A Hysteresis Parameter Identification of Current Transformer Core

ZHOU Shihao, PAN Benren, WAN Yong, WANG Guannan, ZHANG Yan, GUI Xiaozhi   

  1. Electric Power Research Institute of Jiangxi Power Grid Co., Ltd., Nanchang 330096, China
  • Received:2019-08-01 Revised:2020-01-04 Published:2021-07-12
  • Supported by:
    This work is supported by the Science and Technology Project of SGCC (No.521820180017)

摘要: 模拟电流互感器铁心的动态磁化过程,需要建立准确的磁滞模型。改进J-A磁滞模型能够更准确地模拟铁心的磁滞特性。针对改进J-A磁滞模型关键参数难以确定的问题,提出了一种基于烟花算法的电流互感器改进J-A磁滞模型参数识别方法。该方法首先建立改进J-A模型,然后建立参数识别模型,以求出的磁感应强度与已知的磁感应强度之间的绝对误差之和最小为目标函数,将模型参数作为变量引入烟花优化算法中求解。在产生新火花的时候,增加了对火花是否满足约束条件的校验。用所提方法分别识别P型和PR型电流互感器磁滞曲线,结果表明求得的磁滞曲线与实际曲线偏差小,增加的火花校验环节能够明显减少算法收敛所需要的迭代次数,证明了所提方法的有效性和自适应性。

关键词: 电流互感器, 改进J-A模型, 磁滞曲线, 烟花算法, 参数识别

Abstract: To simulate the dynamic magnetization process of the current transformer core, an accurate hysteresis model needs to be established. The improved J-A hysteresis model can simulate the hysteresis characteristics of core more accurately. In this paper, an improved J-A hysteresis model parameter identification method for current transformer based on fireworks algorithm is proposed to determine the key parameter of the improved J-A hysteresis model. Firstly, the modified J-A model is established, and then the parameter identification model is constructed. The sum of the absolute error between the obtained magnetic induction and the known magnetic induction are taken as the objective function, and the model parameters are introduced into the fireworks algorithm as variables. When a new spark is generated, the check of whether the spark meets the constraints is added. Finally, the hysteresis curves of P-type and PR-type current transformers are identified by the proposed method. The results show that the obtained hysteresis curve has a small deviation from the actual curve. The added spark check link can reduce the number of iterations required for algorithm convergence. The effectiveness and adaptability of the proposed method has been verified by simulation results.

Key words: current transformer, modified J-A mode, hysteresis curve, fireworks algorithm, parameter identification