Research On Techniques Of Detection And Estimation Parameters Of Lithium-Ion Battery In Formation Process

Battery formation is an important part of lithium battery production, and determines the performance of the battery. Lithium formation devices using precise detection and parameter estimation techniques, it is possible to obtain exact parameters of lithium battery, and implement accurate control into the process, thus ensuring the quality of the lithium battery. Formation parameters include lithium battery voltage,current, temperature, smoke, capacity, state of charge(State of Charge, SOC) which is estimated based on measured data, SOC is the key parameters in the formation process and plays an important role on the battery performance evaluation. In this paper, some technology-related of parameter detection, and SOC estimation algorithm based on the lithium battery model are researched, which use lithium battery formation system as the hardware platform. The main contents of this paper are as follows:1. These is interference in the battery voltage and current collection process, this paper presents a low-pass filter with a preventing pulse jamming filter combination of programs which can get accurate measurements. In order to make the long-running lithium equipment to maintain a high accuracy, automatic calibration method to detection channels is proposed. On the basis of the measurement accuracy, the closed-loop control algorithm and system of distributed control strategy are researched in this paper.2. Starting from the lithium battery model, the basic properties of the lithium battery are researched firstly. Model parameters are offline identified according the results of experiment. In order to meet online estimation model parameters, recursive least squares method for model parameter identification is used. Specific recursive algorithm of estimation SOC is derived by the extended Kalman filter, unscented Kalman filter, particle filter and unscented particle filter principle. The lithium battery test in working conditions is conducted into the platform of lithium battery formation,and then comparative analysis of the estimated performance of various filtering algorithm is also conducted.3. The sources of SOC estimation error and improved methods are analyzed.According to the time-varying characteristics of lithium model parameters, an online adaptive estimation method is proposed. Firstly, the model resistance is estimated usingthe extended Kalman filter, then the resistance parameter as the SOC estimation algorithm of a known quantity, so that the SOC estimation accuracy is further improved.The lithium battery test in working conditions is conducted, algorithm simulation,comparative analysis of the performance of the adaptive algorithm and non-adaptive algorithms is conducted. In the end, the robustness of the adaptive algorithm is verified.4. The actual total capacity for lithium battery is online estimated according to the information of lithium battery formation process, and then it is used to correct the value of SOC in order to get higher precision in the entire cycle of formation. According to the characteristics of the platform, the SOC adaptive estimation algorithm is embedded in the underlying platform. In order to ensure the stability of the value, algorithm in root form is implemented in the embedded system.