Research On Coordinated Control Strategy Of Hybrid Energy Storage Array System

At present,the development direction of smart grid is green,low-carbon,stable and efficient.The realization of green and low-carbon is bound to need to promote the development and application of renewable energy.Whether it is large-scale centralized access of renewable energy or small-scale distributed access,energy storage technology is required to provide strong functional support to achieve stable and efficient operation of clean energy system.The main problem that restricts the large-scale application of energy storage technology is economy.The emergence of hybrid energy storage technology provides a cost-effective energy storage application solution.Through the research of hybrid energy storage array system(HESAS)control technology,this paper aims to explore the application control problem of hybrid energy storage technology with large capacity and high power,and improve the acceptance ability of power system to intermittent renewable energy.This paper takes battery energy storage,supercapacitor energy storage and HESAS as the research object,and follows the research idea from part to whole.Starting from the control problem of battery energy storage array system with a single energy storage medium,and extending to the HESAS coordinated control problem of different energy storage medium,a series of detailed researches were carried out from shallow to deep.The research results have certain theoretical reference value for the promotion of hybrid energy storage system(HESS).The main research contents of this paper are summarized as follows:Taking the modular battery energy storage system as the research object,the coordinated control of the same energy storage medium is studied.In this chapter,a grouping control strategy suitable for battery energy storage array system is proposed.By adding virtual control variable to the classical consensus model,an improved consensus algorithm based on model predictive control is obtained.And this algorithm has better dynamic and stability.According to the state of charge(SOC)of different cell,the power instruction can be allocated quickly and adaptively.In addition,by dividing the battery energy storage array system into charge group and discharge group,and according to the charge/discharge control strategy developed,the charge/discharge conversion times of the battery unit can be effectively reduced,thus delaying its lifetime degradation process.Finally,the feasibility and superiority of the proposed grouping control strategy are verified by simulating grouping and non-grouping energy storage structures.Based on the research of the control strategy of battery energy storage array system,the power distribution among the same energy storage medium is further studied.Furthermore,it extends to the problem of coordinated control among different energy storage media,and then a two time-scale hierarchical coordinated control method suitable for modular HESAS is proposed.HESAS is a hierarchical and distributed control structure,and the coordinated control layer adopts the two time-scale optimization model based on MPC.The long time-scale optimization model provides power reference value for the short time-scale optimization model to realize the reasonable distribution of high and low frequency power between two energy storage media.And always constrain the SOC of battery energy storage and supercapacitor energy storage in the safe operating range.For the distributed control of the power of the same energy storage medium,the distributed control layer adopts a weighted discrete consensus algorithm based on distributed MPC.It decomposes the power instruction from the coordinated control layer into the power target value of each energy storage unit.Compared with the classical consensus algorithm,the proposed algorithm has fewer iterations and a larger upper bound of step size.Moreover,the partial or total failure of the distributed MPC controller will not affect the final convergence result of the algorithm,so it has good stability and robustness.Finally,the simulation results show that hybrid energy storage technology has great advantages in prolonging battery energy storage life.In the above research process,it is found that to realize the distributed control of multiple energy storage units requires more communication resources.Therefore,the event-triggered theory is introduced to alleviate the problem of too frequent communication in multiple HESSs distributed control.Firstly,the principle and advantages of event-triggered theory are introduced in detail,and the difficulties in its application are discussed.Secondly,a two-level voltage control structure is proposed for the distributed control of DC microgrid with multiple HESSs.Each HESS's local control adopts adaptive droop control method.According to the different SOC of HESS battery and supercapacitor,the droop coefficient can be adjusted in real time.The high/low frequency decoupling of the internal reference current of HESS is realized by a series of first-order low-pass filters behind the voltage control loop,so as to realize the coordinated control goal of steady-state power provided by battery energy storage and instantaneous power provided by supercapacitor energy storage.The upper control is the distributed control layer.It uses a sparse communication network to regulate the average voltage and the proportional current of each HESS to improve the problems arising from a large bus voltage deviation and poor power sharing accuracy caused by virtual and line resistances.Finally,based on the above two-level control structure,an event-triggered controller is introduced,and then two different distributed event-triggered control methods are proposed.The first method is based on the aperiodic sampling method,and through detailed derivation,it is proved that there is no Zeno behavior in this method.The second method is based on the periodic sampling method,which is more compatible with the existing time-triggered control method.By periodically calculating the event-triggered function,the Zeno behavior is essentially avoided.The two event-triggered control methods only need the state error of the local HESS and the last state of the adjacent HESS to detect the event-triggered.By setting a prejudgment threshold of the event-triggered function,the triggered time is further reduced during the steady-state operation of the DC microgrid.The feasibility and effectiveness of the above two methods and two-level voltage control structure were verified by setting up corresponding DC microgrid simulation model in MATLAB/Simulink software.