SOC Estimation and Pack Balance for LiFePO4 Battery Management System

碩士 === 明志科技大學 === 電機工程系碩士班 === 107 === This study developed a battery management system for a LiFePO4 battery pack, which has a SOC estimation method with calibration points and an active battery balance function. Based on the microcontroller, the current of the pack is measured by a current sensor...

Full description

Bibliographic Details
Main Authors: HUANG, YEN-YING, 黃彥穎
Other Authors: WU, CHI-YAO
Format: Others
Language:zh-TW
Published: 2019
Online Access:http://ndltd.ncl.edu.tw/handle/y3c2zp
Description
Summary:碩士 === 明志科技大學 === 電機工程系碩士班 === 107 === This study developed a battery management system for a LiFePO4 battery pack, which has a SOC estimation method with calibration points and an active battery balance function. Based on the microcontroller, the current of the pack is measured by a current sensor and the voltage of each cell is read by a multi-cell monitoring chip. Then the SOC of the pack is estimated by the Ah method with calibration points. Furthermore, the difference in SOC between the cells is estimated and the time and direction required for energy transfer between a cell and the pack are calculated. Finally, by a high-efficiency bidirectional multi-cell chip and bidirectional flyback circuits, the energy was transferred to achieve battery balancing. Numerous SOC estimation methods for a LiFePO4 battery pack have been proposed. The open-circuit voltage method is available that needs a long period of rest, so the method is impractical. The Kalman filter method requires an accurate battery model to estimate the SOC of a pack. However, even the Ah method has starting point errors, but it is more accurate and practical method once the errors can be eliminated in a timely manner. Due to a LiFePO4 pack would be in unbalancing state under long-term charge and discharge and the SOC of the pack in SOC 20% to SOC 80% range is difficult to estimate. Numerous balancing strategies for the pack have been proposed. Among them, the active balance method using a bidirectional flyback DC-to-DC converter is simplest and most efficient. This study developed a battery management system with a SOC estimation function and an active balancing function. In the experiments, four cells were connected in series to form a pack for SOC estimation and active balance experiments. The experimental results show that the initial error and cumulative error of the SOC estimation caused by the Ah method can be significantly eliminated. The test results also confirmed that the balance method used in this study can balance the imbalance of the pack and increase the total capacity of the pack.