The Application of UKF Algorithm for type Lithium Battery SOH Estimation

Transcription

1 Applied Mechanics and Materials Online: ISSN: , Vols , pp doi: / Trans Tech Publications, Switzerland The Application of UKF Algorithm for type Lithium Battery SOH Estimation Zhaoping Chen 1, a, Qiuting Wang* 2,b 1 51 Huzhou Road, Hangzhou, Zhejiang, China 2 51 Huzhou Road, Hangzhou, Zhejiang, China a @qq.com, b @qq.com Keywords: type lithium battery, SOH, UKF, Ohmic resistance Abstract. Lithium battery is widely used in recent years. In this paper, an improved battery model combined with the equivalent circuit model and the electrochemical model is established. The main efforts of our study are: Firstly, the Ohmic resistance of the battery model is identified online based on the Unscented Kalman Filtering (UKF) algorithm. Secondly, the estimation model of the State of Health (SOH) for the type battery is established. Thirdly, an improved battery SOH estimation method based on UKF algorithm is proposed. The experimental results indicate that our new battery model considers the different value of the battery internal resistance on the different working condition, like the different voltage, the different current and the different temperature. Introduction Battery State of Health (SOH) describes the battery performance at the present time compared with the performance at ideal conditions (when the battery was new) [1, 2-6]. It is a measurement that reflects the battery performance and health status. In automotive applications, the estimation of the battery actual SOH is rather crucial to predict the availability of power and energy in hybrid electric vehicles (HEV) and electric vehicles (EV). In theory, by measuring the battery capacity through charging/discharging working periods with the referenced method at certain temperature and the present capacity, the battery SOH can be obtained. Ning G el at. established an empirical model indicating the battery life-cycle, according to the large amount of experimental results. This model is not appropriate for each type of the lithium battery because of the different physical characteristics [7]. Ramadass P et al. presented a mathematical model defining the capacity attenuation of the lithium battery [8]. Alvin J et al. proposed a new battery SOH estimation algorithm based on the fuzzy logic theory [9]. Pillerr G put forward a new method for estimating the change-role of the internal resistances and estimating the real-time capacity of the battery [10]. This method is based on the Kalman filtering algorithm and it indicated a new and more validate method to estimate the battery SOH. Our main efforts are: Firstly, the SOH estimation method for type lithium battery is improved based on UKF algorithm. Secondly, 500 times of the charging/discharging experiments have been done to obtain the enough and reliable data. Thirdly, the function indicating the relationship between the battery internal resistance and the battery life-cycle performance was established. Finally, the battery SOH estimation model was established based on the UKF algorithm and the results. SOH Estimation Method for type Lithium Battery The cycle-life of the lithium battery will be shorter and shorter when the battery is disfunction or sudden failure during the working period. Recently, there are three main SOH estimation methods for the single cell or the battery pack. Two of the methods are used in the practical conditions: The first method is based on the battery internal characteristics [11]. The corresponding function between the cycle-life of the battery and its parameters is established based on the main characteristics during the cell aging process. The second method is based on the data driving theory [12]. According to the advantages and disadvantages of the two methods discussed above, the online-identification method All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of Trans Tech Publications, (ID: , Pennsylvania State University, University Park, USA-19/02/16,07:10:24)

2 1080 Computer and Information Technology for estimating the battery internal-resistances based on UKF algorithm is proposed. The battery SOH can be estimated accurately after the internal resistances are obtained. Figure 1 shows the flowchart of the online SOH estimation method for type lithium battery. Single Battery(18650) Voltage by Battrery Current by Battery Equivalent Circuit Model Voltage by Model Battery Aging Process R 0 UKF SOH Figure 1. The structure of online SOH estimation method Estimation of Battery SOH Based on UKF Figure 2 shows the structure of the equivalent circuit model for type lithium battery. The advantage of this model is that the inconsistent factor during the charging/discharging working V periods is considered. The main parameters of the circuit model are as follows: OC is the open-circuit Rpc voltage (OCV) of the lithium battery and is the function with the battery SOC, is the Rpe R concentration polarization resistance, is the electrochemical polarization resistance, 0 is the Rpc Rpe Ohmic resistance. and are the polarization resistances which indicating the dynamic R characteristics of the battery charging and discharging process, S is the load resistance during the battery discharging working period. Figure 2. The two order circuit equivalent model for lithium battery The Main Algorithm for Battery SOH Estimation Based on UKF I The input vector of the battery model is the circuit terminal-current defined as 0, the output V vector is the different-value (D-value) between the open-circuit voltage defined as O and the V terminal voltage defined as out. The equation in the frequency domain based on Kirchhoff and Laplace transformation theory can be obtained as follows: (1)

3 Applied Mechanics and Materials Vols The differential form of the battery model is: (2) Supposed that and, then the discrete observation equation based on UKF is defined as: (3) The equation indicating the relationship between the parameters of the battery model and the parameters of the function (1) and (2) can be obtained based on the recurrence method, the equation is as follows: (4) According to the experimental results, the DV of the internal resistance can be used as the calculation standard for the degree of the battery capacity decay [13]. The online-estimation equations based on the standard UKF algorithm are defined as follows: r where, k is the DV of the Ohmic resistance and its value changes slowly during the working period. The function (6) describes the output measurement function based on DV of the Ohmic v resistance. k is the estimated error of the battery SOH. The Experimental Results and Its Analysis A. The Experimental Results Figure 3 shows the three different curves of the battery terminal voltage when the lithium battery is working on 0.3 Coulomb (0.3C) discharging current, 1 Coulomb (1C) discharging current and 2 Coulomb (2C) discharging current. Figure 6 indicates that the greater the discharging current is the less time it needs to get to the cut-off voltage. The type lithium battery with high capacity is selected in our study because of its good charging and discharging performance. The main purposes of our study are finding out the main factors affecting the circle-life of the lithium battery and establishing the function indicating the relationship of the battery SOH, the battery open circuit voltage (OCV) and the battery SOC. Figure 4 shows the fitting curve of the OCV-SOC relationship using the least-squares algorithm and the OCV-SOC relationship using the experimental results. After enough experiments and analysis results, the two order function was chosen to be the fitting equation. (5) (6)

4 1082 Computer and Information Technology Figure 3. The discharging voltage curve of different current Figure 4. The fitting curve of the relationship between the battery OCV and the battery SOC B. The Experimental Result and Its Analysis R Figure 5 indicates the online-identification results of the Ohmic resistance 0 using UKF algorithm and the true experimental data obtained by the resistance tester. These experiments were operated on the conditions of the 0.3C discharging current and 2C discharging current. The horizontal ordinate of the figures indicate the battery SOC value. The two longitude coordinates indicate the estimation value of the Ohmic resistance using the UKF algorithm and the measured value using the resistance-tester, respectively. Figure 5 (a) indicates that the value of the Ohmic resistance is between 49 m and 50 m when the battery is working at the current of 0.3C (the small current). Besides, the greater the SOC value is, the larger the estimation error based on the UKF algorithm is. Figure 5 (b) indicates that the value of the Ohmic resistance is between 38 m and 39 m when the battery is working at the current of 2C (the large current). Although the estimation error of the Ohmic resistance is large at the initial moment, the estimation results can convergence to the true value using the UKF algorithm. Finally, the experimental results confirm that the estimation error of the Ohmic resistance is within 5% using our new online-identification method based on UKF algorithm.

5 Applied Mechanics and Materials Vols Figure 5(a). The estimation value and experimental value of the battery Ohmic resistance with 0.3C current Figure 5(b). The estimation value and experimental value of the battery Ohmic resistance with 2C current Conclusion and Future Work Our main effort is proposing a new SOH estimation method for type lithium battery based on UKF algorithm. The main works have been done as follows: Firstly, according to the basic knowledge, the main factors causing the battery capacity attenuation have been analyzed systematically and comprehensively. The experimental and simulation results can be used to provide a theoretical principle for the SOH estimation method. Secondly, after analyzing the traditional methods for the battery SOH estimation problem, a new online-identification method based on UKF algorithm has been proposed. Last but not least, the adaptive model of the battery SOH estimation for the type battery was established based on the UKF algorithm. This model is based on the online-identification results of the battery internal resistances. Acknowledgements This work is funded by the scientific research fund of Zhejiang Provincial Education Department under project no. Y This work is also funded by the Hangzhou Municipal Science and Technology Project under project no. Y B26. This work is also funded by the Zhejiang University City College Teachers Fund under project no J

6 1084 Computer and Information Technology * Corresponding Author:Qiuting Wang received master degree and PH.D degree in Huazhong University of Science and Technology, Wuhan, China. She is currently a teacher in Department of Information & Electrical Engineering, Zhejiang University City College. References [1] Li S. G., Sharkh S. M., Energy and Battery Management of a Plug-In Series Hybrid Electric Vehicle Using Fuzzy Logic, IEEE Transactions on Vehicular Technology, 2011, 60 (8), pp: [2] M. Coleman, W.G. Hurley, L. Chin Kwan, Accurate Prediction of Magnetic Field and Magnetic Forces in Permanent Magnet Motors Using an Analytical Solution, IEEE Transactions on Energy Conversion, 2008, 23 (3), pp: [3] Tucker, J.L., The right tools for the right measurement, IEEE Instrumentation and Measurement Magazine, 2011, 14 (1), pp: [4] B. Saha, K. Goebel, S. Poll, J. Christophersen, Prognostics Methods for Battery Health Monitoring Using a Bayesian Framework, IEEE Transactions on Instrumentation and Measurement, 2009, 58 (2), pp: [5] J. Remmlinger, M. Buchholz, M. Meiler, P. Bernreuter, K. Dietmayer, State-of-health monitoring of lithium-ion batteries in electric vehicles by on-board internal resistance estimation, Journal of Power Sources, 2011, 196 (12), pp: [6] K. Jonghoon, L. Seongjun, B.H. Cho, Complementary Cooperation Algorithm Based on DEKF Combined With Pattern Recognition for SOC/Capacity Estimation and SOH Prediction, IEEE Transactions on Power Electronics, 2012, 27 (1), pp: [7] Ning G, Haran B, Popov B N, Capacity fade study of lithium-ion batteries cycled at high discharge rates, Journal of Power Sources, 2003, 11(7), pp: [8] Ramadass P, Haran B, White R, et al, Mathematical modeling of the capacity fade of Li-ion cells, Journal of Power Sources, 2003, 12(3), pp: [9] Alvin J, Craig F, Singh P, Determination of state-of-charge and state-of-health of batteries by fuzzy logic methodology, Journal of Power Sources, 1999, 80, pp: [10] Plett G, Extended Kalman filtering for battery management systems of LiPB-based HEV battery packs. Part 3, state and parameter estimation, Journal of Power Sources, 2004, 134(2), pp: [11] Gao Fei, Li Jianling, Research progress on lithium-ion power battery life prediction, Electronic Components And Materials, 2009, 28(6), pp: [12] Wei Xuezhe, Xu Wei, Sheng Dan, Internal resistance identification of Li-ion battery and its application in battery life estimation, Power Technology, 2009, 33(3), pp: [13] Cao Jianhua,Gao Dawei, An Experimental Study on the Service Life of LiMn2O4 Battery for Vehicles, Automotive Engineering, 2012, 34(8), pp:

7 Computer and Information Technology / The Application of UKF Algorithm for type Lithium Battery SOH Estimation /