基于CS-PSO算法的电池储能系统多目标优化运行策略

Multi-objective Optimal Operation Scheme Based on CS-PSO Algorithm of Battery Energy Storage System

为了充分发挥储能系统在智能配电网中的积极调节作用,提出了一种统一为成本量纲的电池储能系统多目标优化运行模型。该模型以一个完整调度周期的配电网购电成本、网络损耗费用及电压调节费用均最小为目标函数,以电池储能系统的充/放电功率为控制变量,并确保储能系统在整个调度周期的能量守恒及容量约束。再应用层次分析法计算各子目标权重,化多目标函数为单一综合目标函数。针对所提出的电池储能系统优化运行模型,提出一种改进的混合粒子群优化算法—纵横交叉粒子群优化(CS-PSO)算法。将纵横交叉算子引入粒子群算法,并采用交叉搜索的方法以维护种群多样性,再以电池荷电状态为粒子位置矢量元素,实现完整调度周期内储能系统优化运行策略的求解。最后,对含高渗透率分布式发电单元和电池储能的IEEE34节点算例进行仿真,对比分析了3个单一单目标与本文多目标的优化结果以及3种智能优化算法的计算性能,还分析了储能系统优化运行对系统电压质量的影响。仿真分析结果表明:多目标优化能够充分利用储能系统为配电网提供多种服务,使配电网获得最大综合效益;CS-PSO算法在求解非线性规划问题时具有很好的收敛特性及较高的计算效率,从而验证了所提模型及算法的有效性。

In order to make full use of the battery energy storage systems（BESS） in the smart distribution network（SDN）, a multi-objective optimization model（MOPM） for BESS optimal operation was established, whose objective was to minimize the electricity purchase cost, energy loss fee and voltage regulation cost of the SDN by controlling the charging or discharging power of BESS. And the energy conservation and capacity limits of BESS were included in these constraint conditions of the MOPM. Then an analytic hierarchy process was applied for calculating the weight of each sub-objective function so as to transform the multi-objective function into a single comprehensive objective. To efficiently solve the optimal model, an improved particle swarm optimization algorithm —crisscross and particle swarm optimization algorithm（CS-PSO） was presented by introducing the vertical and horizontal crossover operator into the PSO algorithm and using the method of cross searching to maintain the diversity of the population. Meanwhile, the states of charge of BESS were selected as the elements of particle position vector for solving the model. Finally, the IEEE34-bus system with distributed generations（DGs） and BESS was simulated to demonstrate the effectiveness of the proposed optimal model and algorithm. Compared with the solutions of three single-objective optimal modes, the results showed that the multi-objective optimization was able to make good use of the energy storage system and provide a variety of services for SDN to obtain the maximum comprehensive benefit. Meanwhile, the calculation performances of three intelligent optimization algorithms were compared, which showed that the CS-PSO algorithm had good convergence and computational efficiency in solving the nonlinear programming problems. Furthermore, the influence of the optimal operation of BESS on the system voltage profile was also analyzed, which showed that the multi-objective optimal operation of BESS could improve the voltage quality.