Study On Implementation Of State Of Charge Estimation Method Based On FPGA For Lithium-ion Battery

State of Charge(SOC), one of the most important parameters of battery, is used to characterize the remaining battery electric quantity, which can not be measured directly, and therefore needs to be estimated quickly, accurately and efficiently. In this paper, a fast SOC estimation for lithium-ion battery based on Extended Kalman Filter(EKF) is achieved with an embedded implementation shceme based on Field Programmable Gate Array(FPGA) as the core. Main steps can be listed as follows:The research background and current situation at home and abroad of FPGA scheme to achieve SOC online estimation are introduced. EKF algorithm is choosed to achieve SOC estimation after comparing the advantages and disadvantages of the Battery Management System(BMS) mainstream methods of SOC estimation. By analyzing the implementation methods of various types of hardware and software used in the actual system to estimate SOC, the author selects FPGA as the processor to achieve SOC real-time estimation and introduces the design process of the embedded implementation shceme based on FPGA in detail.The SOC of lithium-ion batteries is estimated via a fast EKF algorithm. The author establishes a first order Thevinan equivalent circuit model of lithium-ion battery and identifies the resistances and capacitances parameters by Recursive Least Square(RLS) method. The accuracy of the first order Thevinan model is validated under custom condition. The author designs an EKF estimator based on fast matrix operation method, estimates the value of lithium-ion batteries’ SOC and also compares the traditional matrix operations method to the fast matrix operations method, which shows that the later requires less time complexity and less running time.An embedded shceme based on FPGA is presented to implement SOC estimation. The author builds the System of Programmable Chip(SOPC) hardware system by using Qsys tool in Quartus II. The SOC estimator based on EKF is designed with C/C++ language under the Nios II software integrated development environment. In the established SOPC embedded system, the author carries out offline validation of EKF algorithm based on traditional matrix operations method and fast matrix operations method, which demonstrate that the EKF algorithm based on fast matrix operations has lower time complexity and needs less time.The author establishes an online experimental test platform for real-time monitoring of battery SOC. A real time data acquisition circuit for the electric current and voltage of A123-26650 Li Fe PO4 is designed, whose collected data of electric current and voltage will be A/D converted by ADS1115 chip. Digital signal after being A/D converted will be transmitted to the EKF estimator in SOPC embedded system to estimate SOC. Through RS232 serial port, the electric current, voltage and the value of SOC will be transmitted to the host computer and be monitored and preserved in real time.The hardware circuit design based on SOC estimator is achieved by the designe of the pure hardware scheme of EKF estimator. Using Verilog Hardware Description Language(HDL), the author establishes a hardware system of EKF estimator, which is tested and validated in the Model Sim. As to the resources occupancy rate and running time of hardware system, the author compares the FPGA pure hardware scheme with the FPGA embedded scheme, which turns out that FPGA pure hardware scheme can improve the operating speed.