Please use this identifier to cite or link to this item:
|Title:||Lithium-ion battery modeling and state-of-charge estimation||Authors:||Tjandra, Rudy||Keywords:||DRNTU::Engineering::Electrical and electronic engineering::Power electronics||Issue Date:||2017||Source:||Tjandra, R. (2017). Lithium-ion battery modeling and state-of-charge estimation. Doctoral thesis, Nanyang Technological University, Singapore.||Abstract:||As there is a growing interest in the electrifying transportation industry and renewable energy in the power system, the accurate battery model has seen increasing relevancy in these highly demanding industries. The battery model is used for the Battery Management System (BMS) to estimate the battery performance during operation (e.g. available power and remaining energy estimation) and for the system simulation study during the design stage (e.g. optimized battery size). To produce an accurate model, it is necessary to understand the battery behaviour under various possible operating conditions, more specifically in relation with the different temperature and the current rate values. As the battery performance degrades over repeated life cycles, the model should account for the age factor as well. In this research work, the Second Order Equivalent Circuit (SOEC) model is chosen over other equivalent circuit models because it offers accurate simulation result and fast computational time. To ensure good accuracy in SOEC model, it is important to extract the parameters accurately. Hence, model parameterization has been developed to fit the voltage response of specific test, in this case, the Hybrid Pulse Power Characterization (HPPC) test. Based on the model parameterization results, battery parameters are found to be dependent on the cycle age, temperature, and the current rate. As the battery ages, its model parameters gradually increase. Depending on its operating temperature, the model parameters have higher parameter values at high temperature than those at low temperature. For the current rate, the model has higher parameter values at a low rather than a high current rate. To ensure an accurate battery model, we need to incorporate these dependencies. In this work, we focus on the Lithium Iron Phosphate (LFP) battery over other chemistries of lithium-ion battery technology. This is on the basis that the LFP battery offers excellent thermal stability (inherent safety), longer cycle life and less expensive material than other commercially-available lithium-ion battery chemistries. However, LFP battery has high hysteresis in Open Circuit Voltage viii (OCV). It complicates the relationship between State-of-Charge (SOC) and OCV as the OCV value is no longer solely dependent on SOC value, but also on its history. Hence, it is necessary to develop a model which is able to account for this phenomenon. In this work, Discrete Preisach model (DPM) is proposed to account for both the current SOC input and its history. Although simulation result shows less accuracy on the constant current and the pulse discharge tests, it shows improved accuracy in the dynamic test which is more relevant to specific applications (such as HEV or ESS in power system). Furthermore, both SOEC model and DPM are utilized in EKF based SOC estimation method. The SOC estimation method has shown good results in the performance and robustness evaluation. We have also investigated the impact of aging on the SOC estimation using aged battery cell test. We compared SOC estimation based on battery model without and with aging consideration. Slight improvement can be noticed in the battery model with aging consideration. Moreover, we also investigated the influence of hysteresis on SOC estimation accuracy. We compared SOC estimation to the battery model with and without hysteresis consideration. The result shows significant maximum error reduction after considering hysteresis in OCV.||URI:||http://hdl.handle.net/10356/72699||DOI:||10.32657/10356/72699||Fulltext Permission:||open||Fulltext Availability:||With Fulltext|
|Appears in Collections:||EEE Theses|
Page view(s) 50215
Updated on Jan 19, 2021
Updated on Jan 19, 2021
Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.