Research On SOC Estimation Of Lithium Iron Phosphate Battery In Wind-Solar Energy Storage System

With the rapid development of world economy, the problem of environmental deterioration is getting worse and worse, at the same time, consisting of wind energy, solar energy and battery energy storage system, the wind-solar energy storage system has drawn worldwide attention. As new energy resources, wind energy and solar energy, are characteristic of irregular in time and space, an accurate estimation on state of charge(SOC) is crucial for the lifetime and utilization of batteries.This paper focuses on batteries in battery management system(BMS),and carries research on SOC estimation on lithium iron phosphate battery. The main points are as follows:First, the paper introduces characteristics and how batteries work. Based on the definition of SOC, it analyzed the influence of current and temperature to the full capacity of battery, and combines voltage characteristic performed in charge and discharge working condition. Considering the battery management system should estimate SOC online and precisely, the extended Kalman filter(EKF) is decided as the method of SOC estimation.Second, the following research is preceded with principle of Kalman filter and workings of lithium battery, and general formulas are set for EKF to meet nonlinear system. This paper mainly studied two models, Thevenin and GNL equivalent circuit model, and the GNL model is eventually used to describe dynamic nature on the basis of voltage variation caused by different two kind of working modes. To estimate model parameters and derive relations between SOC and parameters, hybrid pulse power characteristic(HPPC) test is conducted in battery test system. The EKF algorithm takes Ampere-hour integral as its state equation, and circuit model formula as its observation equation, and then ascertains parameters with the extended Kalman filter recursion formulas and MATLAB simulation software.Last, SOC estimation is fulfilled on experiment platform to simulate working condition of batteries and testify and modify the extended Kalman filter algorithm.Afterwards, we transform these formulas into codes, and apply to the BMS. The results show the modified EKF algorithm has a high precision on SOC estimation on a single lithium battery, and it is also suitable for series connected batteries. It turned out to be of great significance in practice.