Research On Consistency Control Of Battery Energy Storage System For Energy Storage Power Station

With the continuous growth of energy demand,the problem of energy shortage has become a global problem.As an essential energy supply,energy storage technology can improve the quality,stability,and reliability of power grid energy supply and has gradually become the focus of attention.As the primary carrier of energy storage technology,energy storage power plants are still facing some problems,mainly in the inconsistency between battery cells,i.e.,there are differences in capacity,internal resistance,and the voltage between battery cells,resulting in short board effect and loop current loss.charge/discharge multiplier and ambient temperature,which amplify the difference between batteries and will seriously endanger the system’s safety if it is not suppressed.In this regard,it is necessary to monitor the battery state in real time and solve the inconsistency between batteries through equalization technology.This thesis conducts in-depth research on battery state estimation and equalization technology for energy storage power stations.The main contributions are stated as follows:(1)Based on the DP model,research on battery state estimation has been carried out.In terms of battery model parameter identification,to improve the traditional forgetting factor recursive least squares parameter tracking ability and algorithm accuracy,VFFRLS is used to identify battery model parameters online.The value of the forgetting factor is optimized in real-time.In terms of online estimation of the battery SOC,the error caused by linearization when estimating the battery SOC with the EKF,The UKF algorithm is optimized and combined with VFFRLS to achieve a joint estimation of state variables,which improves the estimation accuracy.In terms of online prediction of battery SOP,the response to energy storage,there are a large number of constant power input and output requirements in the application process of the power station.A multi-constraint SOP online prediction algorithm based on VFFRLS-UKF for constant power is proposed to guide the battery energy storage system’s charge/discharge power control.Finally,the relationship between the battery cluster SOC and SOP and the battery cells is explored to provide accurate battery module and battery cluster status information for subsequent battery balancing research.(2)A hierarchical equalization structure is established considering the characteristics of the energy storage power plant,comparing and analyzing different equalization variables with SOC consistency as the equalization target,and proposing an equalization strategy based on the hierarchical structure to achieve inter-cluster and intra-cluster equalization.Aiming at the inconsistency of SOC among battery modules connected in series in a battery cluster,a balancing circuit suitable for the energy storage system of an energy storage power station is designed,and an improved balancing strategy is proposed.By comparing the SOC gating needs of each battery module for the target battery,the SOC variable is introduced in the current loop to make the equalization time of each battery consistent,avoiding the problem of repeated charging and discharging of batteries caused by traditional constant current equalization and improving the efficiency of battery equalization.(3)Based on the 3L-ANPC,complete the energy exchange between the battery,the grid,and the load,and realize the parallel battery cluster in the on-grid and offgrid mode.In the off-grid mode,an improved SOC balance control method based on interconnected droop control is proposed,which comprehensively considers the impact of battery cluster capacity configuration and peak power,dynamically and reasonably allocates power,and realizes battery cluster balance and integration.Provide AC voltage support for the system so that the load can be powered typically.In the grid-connected mode,an improved SOC balance method based on PQ control under the SOP constraint is proposed to ensure the consistency of the charging and discharging states of the battery cluster and provide power for the grid.Sending or absorbing power from the grid.In addition,considering the impact of battery energy storage system failures on consistency control,an equalization strategy based on switch locking and equalization factor dynamic update under failure conditions is proposed,which can maintain the average balance of the remaining battery clusters while realizing the lack of cluster operation,and improve the storage efficiency:system stability and reliability.