• The Transactions of the Korean Institute of Power Electronics
• /
• v.26 no.3
• /
• pp.183-191
• /
• 2021

A battery management system (BMS) provides some functions for ensuring safety and reliability that includes algorithms estimating battery states. Given the changes caused by various operating conditions, the state-of-health (SOH), which represents a figure of merit of the battery's ability to store and deliver energy, becomes challenging to estimate. Machine learning methods can be applied to perform accurate SOH estimation. In this study, we propose a Long-Term Recurrent Convolutional Network (LRCN) that combines the Convolutional Neural Network (CNN) and Long Short-term Memory (LSTM) to extract aging characteristics and learn temporal mechanisms. The dataset collected by the battery aging experiments of NASA PCoE is used to train models. The input dataset used part of the charging profile. The accuracy of the proposed model is compared with the CNN and LSTM models using the k-fold cross-validation technique. The proposed model achieves a low RMSE of 2.21%, which shows higher accuracy than others in SOH estimation.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.23 no.4
• /
• pp.247-255
• /
• 2018

This paper proposes a transformer design of a direct cell-to-cell active cell balancing circuit with a multi-winding transformer for battery management system (BMS) applications. The coupling coefficient of the multi-winding transformer and the output capacitance of MOSFETs significantly affect the balancing current transfer efficiency of the cell balancing operation. During the operation, the multi-winding transformer stores the energy charged in a specific source cell and subsequently transfers this energy to the target cell. However, the leakage inductance of the multi-winding transformer and the output capacitance of the MOSFET induce an abnormal energy transfer to the non-target cells, thereby degrading the transfer efficiency of the balancing current in each cell balancing operation. The impacts of the balancing current transfer efficiency deterioration are analyzed and a transformer design methodology that considers the coupling coefficient is proposed to enhance the transfer efficiency of the balancing current. The efficiency improvements resulting from the selection of an appropriate coupling coefficient are verified by conducting a simulation and experiment with a 1 W prototype cell balancing circuit.

• Journal of Internet Computing and Services
• /
• v.16 no.5
• /
• pp.29-38
• /
• 2015

The Energy Storage System stores electricity for later use. This system can store electricity from legacy electric power systems or renewable energy systems into a battery device when demand is low. When there is high electricity demand, it uses the electricity previously stored and enables efficient energy usage and stable operation of the electric power system. It increases the energy usage efficiency, stabilizes the power supply system, and increases the utilization of renewable energy. The recent increase in the global interest for efficient energy consumption has increased the need for an energy storage system that can satisfy both the consumers' demand for stable power supply and the suppliers' demand for power demand normalization. In general, an energy storage system consists of a Power Conditioning System, a Battery Management System, a battery cell and peripheral devices. The specifications of the subsystems that form the energy storage system are manufacturer dependent. Since the core component interfaces are not standardized, there are difficulties in forming and operating the energy storage system. In this paper, the design of the profile structure for energy storage system and realization of private profiling system for energy storage system is presented. The profiling system accommodates diverse component settings that are manufacturer dependent and information needed for effective operation. The settings and operation information of various PCSs, BMSs, battery cells, and other peripheral device are analyzed to define profile specification and structure. A profile adapter software that can be applied to energy storage system is designed and implemented. The profiles for energy storage system generated by the profile authoring tool consist of a settings profile and operation profile. Setting profile consists of configuration information for energy device what composes energy saving system. To be more specific, setting profile has three parts of category as information for electric control module, sub system, and interface for communication between electric devices. Operation profile includes information in relation to the method in which controls Energy Storage system. The profiles are based on standard XML specification to accommodate future extensions. The profile system has been verified by applying it to an energy storage system and testing charge and discharge operations.

• Proceedings of the KIPE Conference
• /
• 2017.07a
• /
• pp.455-456
• /
• 2017

본 논문에서는 리튬 인산철 배터리($LiFePO_4$)의 내부 파라미터 추출 방법으로 전기화학적 기반인 정전류식 간헐적 적정 테크닉(galvanostatic intermittent titration technique;GITT)을 사용하였다. 배터리 관리 시스템(battery management system;BMS) 알고리즘의 기본적으로 들어가는 충방전 저항을 미세 구간으로 나누어 볼 수 있다. SOC(state-of-charge)에 맞는 저항 성분을 찾을 수 있고, 미소 용량 정보를 알아내어 특정 SOC 구간에서의 LFP 배터리 최적 운용 구간을 알 수 있다.

• Proceedings of the KIPE Conference
• /
• 2012.07a
• /
• pp.299-300
• /
• 2012

본 논문에서는 납 축전지를 사용하는 산업용 전기차량의 SOC(State Of Charge)를 별도의 BMS(Battery Management System)장치 없이 추정하는 방법에 대해 기술한다. SOC를 추정하기 위한 기존의 전통적인 방법들 중 전력을 적산하는 방법(Ampere hour counting)이 널리 사용되는데 이는 장치의 내, 외적인 요인에 의해 발생한 오차가 누적될 수 있다. 배터리의 전압을 측정하여 SOC를 추정하는 OCV(Open Circuit Voltage) 방법은 배터리가 안정 상태에 도달하기까지 충분한 휴지 시간이 필요해 실시간으로 적용하기 힘들다. 이 외에 칼만 필터를 이용하는 방법은 시스템을 정확히 모델링해야 하고 계산이 복잡하다는 단점이 있다. 본 연구에서는 전력을 적산하는 방법을 기본으로 하고 배터리의 전압을 적절히 이용하여 누적되는 오차를 보정하는 방법을 제안한다. 제안한 방법에 대해 시뮬레이션 하고 실제로 산업용 차량인 AC 전동 지게차로 실험하여 그 타당성을 검증 하였다.

• Journal of Power Electronics
• /
• v.17 no.5
• /
• pp.1288-1297
• /
• 2017

The estimation of the remaining useful life (RUL) of lithium-ion (Li-ion) batteries is important for intelligent battery management system (BMS). Data mining technology is becoming increasingly mature, and the RUL estimation of Li-ion batteries based on data-driven prognostics is more accurate with the arrival of the era of big data. However, the support vector machine (SVM), which is applied to predict the RUL of Li-ion batteries, uses the traditional single-radial basis kernel function. This type of classifier has weak generalization ability, and it easily shows the problem of data migration, which results in inaccurate prediction of the RUL of Li-ion batteries. In this study, a novel multi-kernel SVM (MSVM) based on polynomial kernel and radial basis kernel function is proposed. Moreover, the particle swarm optimization algorithm is used to search the kernel parameters, penalty factor, and weight coefficient of the MSVM model. Finally, this paper utilizes the NASA battery dataset to form the observed data sequence for regression prediction. Results show that the improved algorithm not only has better prediction accuracy and stronger generalization ability but also decreases training time and computational complexity.

• Journal of the Semiconductor & Display Technology
• /
• v.7 no.2
• /
• pp.13-17
• /
• 2008

In this paper, in order to get the characteristics of the lithium secondary cell, such as charge and discharge characteristic, temperature characteristic, self-discharge characteristic and the capacity recovery rate etc, we build a thermal model that estimate the impedance of battery by experiment & simulation. In this one-dimensional model, Seven governing equations are made to solve seven variables c, $c_s,\;\Phi_1,\;\Phi_2,\;i_2$, j and T. The thermal model parameters used in this model have been adjusted according to the experimental data measured in the laboratory. The result(Voc, Impedance) of this research can be used in BMS(Battery Management System), so an efficient method of using battery is developed.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.21 no.2
• /
• pp.175-181
• /
• 2016

In the conventional E-bike, a 42 V/10 A Li-ion battery drives a 24 V/10 A BLDC motor via a 6-switch PWM DC/AC inverter. The major problems of the conventional battery-fed motor drive systems are listed as follows. To charge the battery, an external battery charger (adapter) is required, which degrades the portability of E-bike users. In addition, given the high-frequency operation of the motor drive inverter, the switching losses are significant, which degrades the whole power efficiency. High-voltage batteries (42 V) require a complex battery management system (BMS), which degrades the reliability of the battery pack. In this paper, an embedded boost-converter battery charger for E-bikes is proposed. The variable output boost converter, which converts 16.8 V battery voltage to the required variable voltage of the inverter input, can use a low-voltage battery and thus improve the reliability of batteries. By varying the inverter input voltage via boost converter, a DC link voltage control method can be applied to reduce the switching frequency of the inverter, which improves the whole power efficiency. Given that the function of a flyback charger is integrated in the proposed boost converter, the portability of the E-bike user can be maximized by excluding an external adapter. The validity of the proposed circuit will be confirmed by operation mode analysis and simulation. Moreover, experimental results of integrative charger using Li-ion battery and 200 W motor test will be showed with a prototype sample as well.

• Journal of Electrical Engineering and Technology
• /
• v.11 no.6
• /
• pp.1664-1673
• /
• 2016

Lithium rechargeable cells are used in many industrial applications, because they have high energy density and high power density. For an effective use of these lithium cells, it is essential to build a reliable battery management system (BMS). Therefore, the state of charge (SOC) estimation is one of the most important techniques used in the BMS. An appropriate modeling of the battery characteristics and an accurate algorithm to correct the modeling errors in accordance with the simplified model are required for practical SOC estimation. In order to implement these issues, this approach presents the comparative analysis of the SOC estimation performance using equivalent electrical circuit modeling (EECM) and noise suppression technique (NST) in three representative $LiCoO_2/LiFePO_4/LiNiMnCoO_2$ cells extensively applied in electric vehicles (EVs), hybrid electric vehicles (HEVs) and energy storage system (ESS) applications. Depending on the difference between some EECMs according to the number of RC-ladders and NST, the SOC estimation performances based on the extended Kalman filter (EKF) algorithm are compared. Additionally, in order to increase the accuracy of the EECM of the $LiFePO_4$ cell, a minor loop trajectory for proper OCV parameterization is applied to the SOC estimation for the comparison of the performances among the compared to SOC estimation performance.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.13 no.3
• /
• pp.184-196
• /
• 2020

In this paper, a single-bit 2nd-order delta-sigma modulator with the architecture of cascaded-of-integrator feedforward (CIFF) is proposed for precision measurement of current flowing through a secondary cell battery in a battery management system (BMS). The proposed modulator implements two switched capacitor integrators and a single-bit comparator with peripheral circuits such as a non-overlapping clock generator and a bias circuit. The proposed structure is designed to be applied to low-side current sensing method with low common mode input voltage. Using the low-side current measurement method has the advantage of reducing the burden on the circuit design. In addition, the ±30mV input voltage is resolved by the ADC with 15-bit resolution, eliminating the need for an additional programmable gain amplifier (PGA). The proposed a single-bit 2nd-order delta-sigma modulator has been implemented in a 350-nm CMOS process. It achieves 95.46-dB signal-to-noise-and-distortion ratio (SNDR), 96.01-dB spurious-free dynamic range (SFDR), and 15.56-bit effective-number-of-bits (ENOB) with an oversampling ratio (OSR) of 400 for 5-kHz bandwidth. The area and power consumption of the delta-sigma modulator are 670×490 ㎛2 and 414 ㎼, respectively.