储能科学与技术 ›› 2020, Vol. 9 ›› Issue (3): 986-992.doi: 10.19799/j.cnki.2095-4239.2019.0288

• 储能测试与评价 • 上一篇    下一篇

基于动态综合型等效电路模型的动力电池特性分析

封居强1,2(), 司玉文1,2, 伍龙1,2, 黄凯峰1,2, 张星1,2   

  1. 1.淮南师范学院机械与电气工程学院
    2.淮南师范学院轨道交通电源与控制工程技术研究中心,安徽 淮南 232038
  • 收稿日期:2019-12-26 修回日期:2020-01-06 出版日期:2020-05-05 发布日期:2020-05-11
  • 作者简介:封居强(1985—),男,讲师,研究方向为信号检测与估计,E-mail:fjq5060912@126.com
  • 基金资助:
    国家自然基金(61801407);安徽高校自然科学研究项目(KJ2019A0692);淮南师范学院校级项目(2018xj17zd)

Analysis of dynamic battery characteristics based on dynamic synthesis equivalent circuit model

FENG Juqiang1,2(), SI Yuwen1,2, WU Long1,2, HUANG Kaifeng1,2, ZHANG Xing1,2   

  1. 1.College of Mechanical and Electrical Engineering, Huainan Normal University
    2.Rail Transit Power Supply and Control Engineering Technology Research Center, Huainan Normal University, Huainan 232038, An’hui, China
  • Received:2019-12-26 Revised:2020-01-06 Online:2020-05-05 Published:2020-05-11

摘要:

针对动力电池模型的内部参数随负载和工况的变化而变化,本文综合考虑电池容量、温度、倍率、迟滞、自放电等因素,设计综合性实验。通过分析动力电池端电压与欧姆极化、电化学极化、浓差极化、不平衡电势、迟滞特性等因素之间的关系,确定动力电池的内部状态变量,提出动态综合型等效电路模型(DS-ECM),借助混合脉冲功率特性法(HPPC)进行动态辨识。针对变化的工况,与Rint、PNGV、二阶RC模型进行比较,并借用赤池信息准则(AIC)衡量模型的优良性。结果表明:本文模型估计的端电压较其它模型平均误差最小降低了52.3%,且AIC值和计算量相对适中,是精度和结构复杂度最佳平衡模型。本研究有助于推动等效电路模型在实际工程中的应用。

关键词: 动力电池, 等效电路模型, 迟滞效应, 赤池信息准则

Abstract:

To determine the internal parameters of the power battery model under changing load and working conditions, this work designs comprehensive experiments considering the battery capacity, temperature, multiplier, hysteresis, self-discharge, and other factors into comprehensive consideration. By analyzing the relationship between the terminal voltage of the power battery and factors such as ohmic, electrochemical, and concentration polarizations; unbalanced potential; and the hysteresis characteristics, the internal state variables of the power battery were determined and a dynamic integrated equivalent-circuit model was proposed. Dynamic identification was performed using a hybrid pulse-power characteristic method. The performance of the proposed model under varying operating conditions was compared with those of the Rint model, the partnership for a new generation of vehicle model, and the second-order RC model. The performance index was the Akaike information criterion (AIC). The average error of the proposed model was 52.3% lower than that of the other models. Moreover, the AIC and calculation values were moderate, implying that the proposed method better balances accuracy and structural complexity than the comparison methods. This study will be helpful in promoting the application of equivalent circuit models in practical engineering.

Key words: power battery, equivalent circuit model, hysteresis effect, Akaike information criterion

中图分类号: