本文選題：鋰離子電池 + 等效電路模型��； 參考：《上海海洋大學(xué)》2017年碩士論文

【摘要】：近年來,隨著經(jīng)濟(jì)發(fā)展,環(huán)境污染的日益加劇,傳統(tǒng)的化石能源已不能滿足人類生產(chǎn)生活的需要。為此,許多交通工具如火車,汽車等開始使用電力作為能源。電力能源可以儲存在多種介質(zhì)中,日產(chǎn)生活中應(yīng)用最廣泛的就是電池。電池的荷電狀態(tài)(SOC)估算是對電池應(yīng)用研究的一個基礎(chǔ),電池在使用中的深度充放電,會大大減少其使用年限,精確的SOC估計可以避免這種情況;剩余電量的準(zhǔn)確顯示,還能幫助汽車控制系統(tǒng)計算可行駛里程,駕駛者也可以更好規(guī)劃行駛路線。以上優(yōu)點(diǎn)說明SOC的研究具有重要的意義。本文從鋰離子電池的化學(xué)結(jié)構(gòu)入手,對鋰電池的原理進(jìn)行了分析,通過階梯充放電實(shí)驗(yàn)研究了鋰電池的充放電電壓特性,分析了實(shí)驗(yàn)環(huán)境溫度、充放電倍率對鋰電池性能的改變。本文針對現(xiàn)有的四種經(jīng)典鋰離子電池電路模型,即內(nèi)阻,Thevenin,PNGV和GNL模型所存在的不足,提出本文所采用的二階RC等效電路模型,并推導(dǎo)出了此二階模型的狀態(tài)空間模型。接著,基于恒流放電實(shí)驗(yàn)數(shù)據(jù)用五次多項(xiàng)式極大逼近了鋰電池的OCV-SOC關(guān)系。然后基于實(shí)驗(yàn)數(shù)據(jù),采用參數(shù)擬合的方法和具有遺忘因子的RLS算法對鋰離子電池二階模型的參數(shù)進(jìn)行了離線和在線辨識。MATLAB仿真實(shí)驗(yàn)研究表明,本文所建立的二階等效電路模型能較好地體現(xiàn)鋰電池的端電壓特性,所建模型精度較高。從而為鋰離子電池的SOC估計奠定基礎(chǔ)。鋰電池系統(tǒng)是一個復(fù)雜的系統(tǒng),考慮到擴(kuò)展卡爾曼濾波(EKF)算法在解決非線性系統(tǒng)濾波問題方面的優(yōu)勢,本文采用EKF對鋰電池的SOC進(jìn)行預(yù)測。首先基于鋰電池的二階等效電路模型,推導(dǎo)出了其線性離散化狀態(tài)空間模型,確定了鋰電池的SOC、極化電壓、噪聲方差陣等的初始值后,采用EKF逐次對恒流放電狀態(tài)、周期性脈沖放電狀態(tài)、FUDS狀態(tài)和BJDST狀態(tài)下鋰電池的SOC進(jìn)行了估計。并將EKF在各工況下的SOC估計結(jié)果與實(shí)際結(jié)果實(shí)施了比較,仿真實(shí)驗(yàn)得出結(jié)論,對于FUDS和BJDST工況,雖然在初期階段SOC的評估誤差較大,但最終EKF都可以順利地預(yù)測鋰電池的SOC改變。這說明本文采用EKF對鋰電池的SOC進(jìn)行預(yù)估是完全可行的,EKF在鋰電池的SOC評估方面具有良好的魯棒性。
[Abstract]:In recent years, with the development of economy and environmental pollution, the traditional fossil energy can not meet the needs of human production and life. To this end, many vehicles such as trains, cars and so on began to use electricity as an energy source. Electric energy can be stored in a variety of media, Nissan life the most widely used is the battery. The state of charge (SOC) estimation of the battery is a basis for the study of battery application. The depth charge and discharge of the battery in use will greatly reduce its service life, which can be avoided by accurate SOC estimation. It also helps car control systems calculate mileage and better route planning. The above advantages show that the study of SOC is of great significance. Based on the chemical structure of lithium ion battery, the principle of lithium battery is analyzed in this paper. The charge-discharge voltage characteristic of lithium battery is studied by step charge-discharge experiment, and the temperature of experimental environment is analyzed. The change of charge / discharge ratio on the performance of lithium battery. In this paper, the second order RC equivalent circuit model is proposed and the state space model of this second order model is derived, aiming at the shortcomings of the existing four classical lithium ion battery circuit models, namely, the internal resistance Theveninn PNGV model and the GNL model. Then, based on the constant current discharge experimental data, the OCV-SOC relation of lithium battery is approximated by the fifth order polynomial. Then, based on the experimental data, the parameter fitting method and the RLS algorithm with forgetting factor are used to study the off-line and on-line identification of the parameters of the second-order model of Li-ion battery. MATLAB simulation results show that, The second order equivalent circuit model established in this paper can well reflect the terminal voltage characteristics of lithium battery, and the precision of the model is high. Thus, it lays a foundation for the SOC estimation of lithium ion batteries. Lithium battery system is a complex system. Considering the advantage of extended Kalman filter (EKF) algorithm in solving nonlinear system filtering problem, EKF is used to predict the SOC of lithium battery. Firstly, based on the second order equivalent circuit model of lithium battery, the linear discrete state space model is derived, and the initial values of SOC, polarization voltage and noise variance matrix of lithium battery are determined, then the constant current discharge state of lithium battery is analyzed by EKF step by step. The SOC of lithium battery under periodic pulse discharge state and BJDST state are estimated. The SOC estimation results of EKF under various operating conditions are compared with the actual results. The simulation results show that for the FUDS and BJDST conditions, the evaluation error of SOC in the initial stage is large. But in the end, EKF can predict the change of SOC of lithium battery. This shows that it is feasible to predict the SOC of lithium battery by using EKF in this paper. EKF has good robustness in evaluating SOC of lithium battery.
【學(xué)位授予單位】：上海海洋大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TM912

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 李百華;郭燦彬;鐘其水;涂濤;;電動汽車鋰電池戴維南等效電路模型參數(shù)辨識研究[J];微型機(jī)與應(yīng)用;2017年01期

2 鄧曄;胡越黎;滕華強(qiáng);;鋰電池開路電壓的預(yù)估及SOC估算[J];儀表技術(shù);2015年02期

3 鄧濤;孫歡;;鋰離子電池新型SOC安時積分實(shí)時估算方法[J];重慶理工大學(xué)學(xué)報(自然科學(xué));2015年01期

4 陳嵐;劉皓U，

本文編號：1954221