快速平均一致与多采样率下储能系统精准功率分配

Precise power allocation with fast average consensus and multi-rate sampling for energy storage systems

针对智能电网中储能系统分布式功率分配问题,本文提出融合了快速平均一致与多采样率的离散时间控制策略,使各储能电池以相同相对剩余电量状态(SoC)变化率进行充放电.为了仅用局部通信得到计算功率分配值需要的全局平均值,在储能电池拓扑结构已知和未知两种情况下,本文分别采用了有限时间平均一致算法和具有最优收敛率的平均一致算法.在储能电池单个控制周期内,平均一致算法以小于电池控制周期的采样周期进行多步迭代,从而得到精确的估计值,实现各储能电池的精准功率分配.不同于传统连续时间控制方法仅给出平均一致算法的收敛性分析,本文不仅建立了单个控制周期内快速平均一致算法的收敛率表达式,而且给出了包含储能电池动力学和平均一致算法的整个控制系统的渐近稳定性分析.所提方法能够更快速、更精准地实现基于相同相对SoC变化率的功率分配.最后,本文通过多组仿真实验验证了所提方法的有效性.

Considering the power allocation of energy storage systems in smart grid,this paper proposes a discretetime distributed control strategy that combines fast average consensus and multi-rate sampling control.The batteries are discharged or charged with the same relative state-of-charge(SoC)variation rate.To obtain the global average value which is utilized to compute the power allocation,under the cases that the battery topology is known and unknown,a finite-time average consensus algorithm and an average consensus algorithm with optimal convergence rate are adopted,respectively.In each control period of the batteries,the average consensus algorithm with smaller sampling period iterates multiple times to obtain accurate estimation value.Different from traditional continuous-time control methods that only provide the convergence analysis of average consensus algorithms,this paper not only presents the convergence rate of fast average consensus in each control period,but also analyzes the asymptotic stability of the control system which contains the battery dynamics and the average consensus algorithm.The proposed method can achieve precise power allocation based on the same relative SoC variation rate.Finally,simulation examples are given to verify the effectiveness of the proposed method.