Lithium Battery SOC Estimation And Marine Battery Management System Design

With the strategy of "Strong Ocean State" and the arrival of the 5G era,the government has paid great attention to unmanned ships,and unmanned material throwing vessels,as new intelligent aquaculture equipment,has become a research hot spot in China.The power lithium battery as the power equipment of unmanned material throwing ship plays a decisive role in the power performance of material throwing ship.The State of Charge(SOC)of lithium batteries is one of the key parameters affecting the efficiency and safety of the power pack,and it is still difficult to estimate the SOC accurately.The focus of this paper is to investigate the SOC estimation of Li-ion battery based on the drive model and equivalent circuit model,and then complete the design of a marine48V50 AH iron phosphate small capacity power Li-ion battery management system based on the actual project of unmanned throwing ship.The specific research work is as follows.For the SOC estimation of lithium batteries based on the data-driven model,the SOC estimation method based on an improved genetic algorithm optimized BP neural network(IGA-BP)is proposed.Firstly,the cosine function is introduced to design the adaptive crossover operator and variation operator to improve the population diversity;then the simulated annealing method’s probabilistic burst jump capability is utilized to prevent the genetic algorithm from discovering the local optimal solution;The neural network’s initial weights and thresholds are then optimized;three working condition data sets,DST,US06,and FUDS,are set up to validate and analyze the improved algorithm by using the University of Maryland lithium battery discharge data.For the estimation of Li-ion battery SOC based on an equivalent circuit,the secondorder RC equivalent circuit model is established,and the online parameter identification is performed using the recursive least squares method with the forgetting factor.Considering that the extended Kalman is affected by temperature,battery aging,and other factors in estimating SOC,the disturbance noise will change in real-time,and the EKF will generate higher-order loss error in the process of linearization,which affects the accuracy of the Li-ion battery charge state estimation,therefore,iterative theory,LM method,and adaptive filtering are introduced for improvement,and the combination FFRLS-AIEKF algorithm-based SOC estimate technique is suggested.Under UDDS operating settings,the revised algorithm’s performance is confirmed.For the 48V50 AH marine lithium battery management system,the STM32F103RBT6 as the main controller,the battery management system was designed based on the BQ76940 power management chip.Visual Studio 2017 software was used to develop the lithium battery monitoring management software on the PC side.It realizes the monitoring of the total voltage,single voltage,current,and temperature of Marine lithium battery packs,and realizes the functions of SOC estimation,battery balance,and charge and discharge control.The battery management system is tested,and the results show that the data collected by the battery management system has little difference from the actual instrument measurement data,and has high accuracy,which can meet the requirements of Marine lithium battery information status monitoring and realize battery balance,and provides a hardware and software design scheme for further realizing the battery management system based on complex SOC estimation algorithm.