随着高渗透率分布式新能源接入电网，其出力的随机性与波动性对配电网的灵活性需求提出了更高的要求。为此提出一种计及智能软开关(soft open point, SOP)与智能负载(smart load, SL)互补特性的配电网灵活性提升方法。首先，对SOP、SL与可控分布式电源（distributed generation, DG）的灵活性特性与数学模型进行分析，从节点灵活性适应度与含SOP线路的负载裕度两个方面建立灵活性评估指标。其次，构建考虑运行成本、电压质量与灵活性指标的目标函数，力求在最小化灵活性缺额与电压越限的情况下实现经济性最优。同时提出了改进鲸鱼优化算法(reconstruction whale optimization algorithm, Re-WOA)对优化模型求解；在该算法中引入了两种改进的搜索捕食策略，并以SL容量裕度指标平衡搜索阶段的SL负载均衡度。最后，通过改进的IEEE 33节点系统验证了该模型对配电网的优化作用与所提Re-WOA算法的有效性。

The integration of new high-permeability distributed energy into the power grid has increased the demand for flexibility in the distribution network owing to the randomness and volatility of its output. A method for improving the flexibility of distribution networks that considers the complementary characteristics of soft open point (SOP) and smart load (SL) is proposed. First, the flexibility characteristics and mathematical models of the SOP, SL, controllable DG, and flexibility evaluation indicators were established from two aspects: node flexibility adaptability and the load margin of lines containing the SOP. Second, we constructed an objective function that considers the operating costs, voltage quality, and flexibility indicators, striving to achieve optimal economic performance while minimizing flexibility gaps and voltages exceeding the limits. Simultaneously, a reconstruction whale optimization algorithm (Re-WOA) was proposed to solve the optimization model, and two improved search predator strategies were introduced in the algorithm. The SL load balancing degree during the search phase was balanced using the SL capacity margin index. Finally, the optimization effect of the model on the distribution network and the effectiveness of the proposed Re-WOA were verified using an improved IEEE33 node system.

随着分布式电源渗透率不断提高,配电网运行面临着源荷不确定性带来的灵活性不足问题。合理配置储能(energy storage system, ESS)和智能软开关(soft open point, SOP)等灵活性资源是解决该问题的关键。为此,从灵活性供需匹配度和传输能力2个方面提出了考虑配电网灵活性提升的ESS和SOP三层协同规划模型。其中,上层以ESS投资成本最小为目标对蓄电池和超级电容器等多时间尺度储能进行规划,中层以配电网年综合成本最小为目标对SOP进行规划,下层以各场景下的网络灵活性最优为目标进行优化,并采用自适应粒子群算法和二阶锥规划算法对模型进行求解。最后,在IEEE 33节点系统上验证了所提方法的有效性。结果表明,相比仅考虑单一时间尺度的规划方法,所提的ESS和SOP协同规划方法的配电网年综合成本降低了4.59%,实现了配电网经济灵活运行。

With the increasing penetration rates of distributed generators, the operation of distribution networks faces the challenge of insufficient flexibility owing to uncertainties in supply and demand. The key to addressing this issue lies in properly allocating flexible resources, such as energy storage systems (ESS) and soft open points (SOP). To this end, a three-level coordinated planning model for ESS and SOP that considers the flexibility of the distribution network is proposed in terms of flexibility, supply-demand matching, and transmission capacity. Within this model, the upper-level plans for multi-timescale energy storage, such as storage batteries and supercapacitors, aim to minimize the ESS investment costs. The middle-level plans for the SOP are intended to minimize the annual comprehensive operation cost of the distribution network. The lower-level plans optimize the objective of network flexibility in each scenario and solve the model using an adaptive particle swarm algorithm and a second-order cone planning algorithm. Finally, the effectiveness of the method is verified using an IEEE 33-bus system. The results show that, compared with the planning method that only considers a single time scale, the coordinated planning approach of the ESS and SOP proposed in this study reduces the annual comprehensive cost of the distribution network by 4.59%, thereby achieving the economic and flexible operation of the distribution network.

多类型分布式电源接入配电网对系统灵活性要求越来越高。基于交直流混合配电网的高灵活性和高可控性,提出了一种考虑灵活性的交直流混合配电网储能双层规划方法。首先,分析系统灵活性需求与灵活性资源运行特性,建立系统灵活性评估指标;然后,采用K-means聚类算法建立配电网中分布式电源和负荷的典型运行场景集合;其次,建立考虑灵活性的交直流混合配电网储能双层规划模型,上层以其年综合成本最小为优化目标,下层以其综合运行成本最少和系统运行灵活性最优为目标,采用非线性化主成分分析法,并结合遗传算法求解规划模型。最后,基于改进IEEE 33节点配网系统验证双层规划模型和储能最优配置方案的合理性与有效性。

The AC/DC hybrid distribution network has the characteristics of high flexibility and high controllability, and will be widely used in the future distribution network. A two-level planning method for energy storage in AC/DC hybrid distribution network considering flexibility is proposed in this paper. Firstly, this paper analyzes the flexibility requirements in the AC/DC hybrid distribution network and the operating characteristics of the flexible resources considered, and then establishes a quantitative calculation index to measure the flexibility in the operation of the distribution network. Then, K-means clustering algorithm is used to establish the collection of typical operation scenarios for the output and daily load of distributed power generation in distribution network. Secondly, a two-level planning model for energy storage in AC/DC hybrid distribution network is established, which takes into account flexibility. The goal is to minimize the comprehensive operation cost of the distribution network and the optimal level of system flexibility, and the genetic algorithm is used to solve the optimal planning scheme. Finally, on the basis of the improved IEEE 33-node distribution network system, the rationality and effectiveness of the two-level planning model and the optimal configuration of energy storage are verified.

在高比例分布式光伏接入的背景下,给配电网运行的经济性与安全性带来了新的挑战。建立了考虑开关重构与需求响应的配电网分布鲁棒优化运行模型。首先,以配电网每日运行成本最小为优化目标,同时考虑开关重构与价格型需求响应,得到配电网拓扑结构优化方案与运行方案。其次,利用光伏历史出力数据,通过聚类获取典型场景与其概率分布,采用1-范数与∞-范数约束不确定性概率分布置信区间,建立数据驱动的分布鲁棒两阶段优化模型。最后,通过修改的IEEE 33节点系统,基于列与约束生产算法(column-and-constraint generation, CCG)利用Cplex求解器进行求解,验证了模型的可行性。

Under the connection of high-penetration distributed photovoltaic generation, new challenges face the economic and safe operation of distribution networks. In this study, a distributionally robust optimization model considering switch reconfiguration and demand response was established to optimize the operation of a distribution network. First, considering switch reconfiguration and price-based demand response, topology optimization and operation schedules of the distribution network were obtained with the optimization objective of minimizing daily operating cost. Second, using the historical photovoltaic (PV) output data, the typical scenarios of PV output and its probability distribution are obtained through clustering, and the 1-norm and ∞-norm are used to constrain the confidence interval of the uncertainty probability distribution, and a data-driven distributionally robust two-stage optimization model is established. Finally, with the modified IEEE 33-node system, the Cplex solver was used based on the column-and-constraint generation algorithm, and the feasibility of the model was verified.