| Single-Flow Zinc-Nickle Battery is a new type of electrochemical energy storage device with simple structure,strong cycling performance,large scale energy storage and low cost.It has got much attention in the field of large-scale energy storage,with high research value and good application prospects.State of charge(SOC)is a key parameters to describe the operating state of a battery,which indicates the percentage of remaining charge,and its accurate estimation is a prerequisite for safe battery use and energy management.In this paper,we focus on the SOC estimation method of single-fluid flow zinc-nickel battery,and conducts basic research on the battery with the goal of improving the SOC estimation accuracy and convergence.Battery modeling is a prerequisite for SOC estimation.In this paper,the advantages and disadvantages of various battery modeling methods were discussed,and the second-order RC equivalent circuit model,which is less complex,was chosen to model the single-liquid flow zinc-nickel battery under the premise of ensuring the accuracy of the model.Secondly,Xinwei battery tester BTS-5V200 A was used to pulse charge and discharge the battery to obtain voltage and current data to establish the open circuit voltage curve,and on the basis of this,curve fitting was used to identify the battery model parameters offline.Finally,the corresponding battery equivalent model was built in MATLAB/Simulink environment,and the validity of the built model and the accuracy of the identified parameters were verified by simulation.SOC estimation is a core functions of the battery management system,and it is of great importance to achieve accurate SOC estimation.This paper analyzed various common battery SOC estimation methods,and the Dual Extend Kalman Filter(DEKF)algorithm with dual lines for multiple time scales was selected after comprehensive consideration.The main feature of this algorithm is that the SOC is estimated while the model parameters are tracked in real time,which can reduce the impact of the time-varying cell model parameters on the SOC estimation.In addition,the estimation ability of Kalman filter algorithm depends heavily on the noise statistical characteristics(process noise covariance,measurement noise covariance)and the initial value of state error covariance,but it is usually selected empirically and hard to get the optimal value,so this paper introduced an improved Archimedes optimization algorithm to optimize the parameters of DEKF algorithm to better perform its estimation performance.The results showed that the parameter-optimized DEKF algorithm has higher estimation accuracy and better convergence and robustness in estimating the SOC of single-fluid flow zinc-nickel batteries.In order to test the practicality of the above SOC estimation method,in this paper,a single-fluid flow zinc-nickel battery SOC estimation platform based on d SPACE DS1103 was designed.The platform is equipped with a charge/discharge module,which can realize the constant-current charge/discharge of single-liquid-flow zinc-nickel batteries,and at the same time send the collected battery terminal voltage,charge/discharge current data and SOC estimation results to the Control Desk of the host computer software for real-time monitoring.The experimental results indicated that the platform can effectively estimate the SOC of single-liquid flow zinc-nickel battery,and the estimation error is within 1%,and the initial value can be converged to the accurate value quickly when the initial value is wrong. |
PDF Full Text Request