http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
테스트벤치를 이용한 배터리 모듈을 위한 BMS 알고리즘 및 펌웨어 개발
박형욱(Hyeongwook Park),윤재중(Jaejung Yun) 한국자동차공학회 2023 한국자동차공학회 학술대회 및 전시회 Vol.2023 No.11
The core technologies of BMS include battery monitoring, diagnosis, control, communication, and battery state estimation. In this paper, the algorithm and firmware of the BMS with the Master-Slave structure were developed using a test bench. The battery voltage, current, and temperature were monitored through the Slave BMS using the Analog Front End (AFE). SPI and UART communication were used to transmit battery data among Slave BMS, Master BMS, and host PC. Master BMS has a micro-controller with calculation and interrupt functions and performs SOC estimation using battery data sensed by the Slave BMS. In addition, this paper presents firmware for Master-Slave BMS using MCU (STM32F746G-DISCO from STM) and AFE (DC2259A from ADI). The measurement errors for voltage, current, and temperature in the battery monitoring results are within 10mV, 10mA, and 2℃, respectively. For the SOC estimation algorithm, an enhanced Coulomb counting method has been proposed, and a real-time SOC estimation system with an error of less than 2% has been implemented.
수동 셀 밸런싱 알고리즘 비교 분석을 통한 밸런싱 최적화에 관한 연구
이재형(Jaehyung Lee),윤재중(Jaejung Yun) 한국자동차공학회 2023 한국자동차공학회 학술대회 및 전시회 Vol.2023 No.11
The passive cell balancing techniques currently commercialized in electric vehicles and railway vehicles include algorithms based on minimum and average voltage values and smart balancing algorithms proposed by TI. The balancing speed and efficiency of these techniques differ depending on the battery type and application, and optimization is required to improve performance and stability depending on the requirements of the system. Therefore, this paper constructed a BMS test bench with a Mater-Slave structure using an LFP battery module to optimize the balancing algorithm. The balancing operation range for experiments of the each algorithm was set between OCV at 80% SOC and full charge voltage. The experimental results of the minimum voltage based algorithm showed that the balancing efficiency was 86.7% and the charging speed was slow. The results of the average voltage based algorithm showed that balancing efficiency was 90% and the charging speed was fast. The smart balancing algorithm had the fastest charging speed but the balancing efficiency was low at 85.3%. Based on the results of each balancing algorithm, we researched to optimize the balancing method to improve the performance and efficiency of a battery system.
시차열전압전류법 분석 및 LSTM 기반의 리튬이온 배터리 SOH 추정
최연호(Yeonho Choi),윤재중(Jaejung Yun) 한국자동차공학회 2023 한국자동차공학회 학술대회 및 전시회 Vol.2023 No.11
Accurate state of health estimation(SOH) provides important indicators to ensure reliability and stability in battery-using applications. In this paper, battery degradation features were extracted through differential thermal voltammetry(DTV) analysis. Through Pearson correlation analysis, SOH was estimated using degradation features with a high correlation with SOH as input to the LSTM model. As a result of the estimation, it was confirmed that DTV-based degradation features have a high correlation with SOH and that high accuracy of SOH estimation can be achieved when estimating SOH using the LSTM model.
오프라인 식별 기반 적응형 ECM 파라미터 기법들의 확장 칼만 필터를 SOC 추정 성능 비교
최연호(Yeonho Choi),박형욱(Hyeongwook Park),이재형(Jaehyung Lee),윤재중(Jaejung Yun) 한국자동차공학회 2023 한국자동차공학회 부문종합 학술대회 Vol.2023 No.5
The extended kalman filter based on the battery model is widely used as an SOC estimation algorithm. However, the conventional extended kalman filter uses a single parameter without reflecting the parameters of the ECM that change according to the SOC. However, if the variation of the ECM parameters are not updated, the SOC estimation accuracy is adversely impact. To improve this problem, many research on adaptive ECM parameters have been conducted. In this paper, among the methods, grouping and curve fitting methods for ECM parameters were compared. Using the battery FUDS profile, the SOC estimation performance and algorithm running time of the conventional method and the extended kalman filter to which the adaptive ECM parameters were applied were claculated through simulation.
LTO 배터리 모듈을 위한 BMS 시뮬레이션 모델 개발
손성재(Sungjae Son),이재형(Jaehyung Lee),박형욱(Hyeongwook Park),윤재중(Jaejung Yun) 한국자동차공학회 2023 한국자동차공학회 부문종합 학술대회 Vol.2023 No.5
Lithium titanium oxide LTO (Li4Ti5O12) batteries are used in applications requiring high capacity and high output, such as eco-friendly railway vehicle. Until now, simulation models of the BMS(Battery Management System) for Li-ion batteries have been developed. However, no research has been conducted on BMS simulation models for LTO batteries. This paper presents a BMS simulation model for a module with 11 LTO batteries connected in series. It was developed in the MATLAB/SIMULINK environment based on a circuit using ADI’s AFE(Analog Front End) IC and LTO-2717397 battery. In the BMS simulation, the SOC estimation and passive balancing algorithms for a LTO battery module were verified.