A Smart Real-Time State of Charge Estimaion Scheme for Battery Energy Storage Systems

Autor: Po-Wen Chen, 陳柏文
Rok vydání: 2016
Druh dokumentu: 學位論文 ; thesis
Popis: 104
In recent years, many countries in the world are actively developing renewable energy based generations (REBG) and smart grids in order to achieve the purpose of reducing carbon emissions. Renewable energy mainly includes wind, solar, hydro and geothermal energies, however REBG affected by the environmental conditions can not provide a stable power. To solve this problem, the battery energy storage system (BESS) can provide a promising solution in smoothing power fluctuations of REBG and maintaining a stable and continuous supply of electricity. In order to provide functionalities according to a given plan, one needs to obtain the battery status in BESS and base on that the corresponding control and management strategies can be determined. Therefore, the real-time estimation of the state of charge (SOC) is a significant issue. There are a number of factors that could affect the SOC of battery, e.g., the charge and discharge currents, ambient temperature, electrolyte concentration, internal impedance and the cycles used may have the effect of aging. In fact, all of these have handicapped the existing methods from being able to accurately estimate SOC of batteries. To compensate the gap between design algorithms in the above methods and consider the feasibility of practical applications, this research proposes a smart real-time SOC estimation method and the lead-acid batteries are chosen for the study. Since the charging or discharging process of the battery has a nonlinear feature, the neural networks (NN) and adaptive neural fuzzy inference system (ANFIS) are used in this research to effectively learn nonlinear characteristics of batteries on-line, whereby the real-time SOC detector is developed. In this work, the Coulomb charge accumulation method is also used to provide a comparative analysis in terms of estimation performances. Several simulation and experimental cases are carried out to demonstrate the feasibility and effectiveness of the proposed real-time SOC detection algorithm.
Databáze: Networked Digital Library of Theses & Dissertations
Externí odkaz: