【目的】针对高比例光伏接入引发的配电网电压波动与网络损耗增加问题，本文提出了一种基于时序综合电压-有功灵敏度的储能电源优化配置策略。【方法】首先，建立节点电压与有功功率变化之间的静态综合灵敏度矩阵，以量化功率注入对所有节点电压偏差的影响。其次，为反映光伏出力与负荷波动的时序特征，引入时序权重因子修正不同时段电压偏移的影响，从而形成时序综合灵敏度指标，用于确定储能系统的优先接入位置。在此基础上，构建同时最小化储能系统综合运行成本、配电网电压偏移和总网络损耗的多目标优化模型。然后，采用改进粒子群优化算法求解多目标优化模型，通过动态调整惯性权重和学习因子，提高算法的收敛速度并避免陷入局部最优。最后，采用生成对抗网络生成多种配电网运行场景，以丰富优化数据集，提高光伏与负荷不确定性下解的鲁棒性。【结果】算例结果表明，与传统电压灵敏度分析方法相比，所提时序灵敏度分析方法有效提升了节点电压调节的针对性和精度。优化后的储能系统配置在确保节点电压保持在允许运行范围内的同时，实现了经济性与可再生能源消纳能力的协同提升。【结论】所提策略显著加快了配电网储能规划过程，提升了其在光伏波动条件下维持电压稳定的能力，促进了高比例光伏配电网的经济与可靠运行。

[Objective] To address the issues of voltage fluctuations and increased network losses in distribution network caused by high photovoltaic penetration, this paper proposes an optimal configuration strategy of energy storage power source based on time-series comprehensive voltage-active power sensitivity. [Methods] Firstly, a static comprehensive sensitivity matrix between nodal voltage and active power variation was established to quantify the impact of power injection on voltage deviation across all nodes. Then, to reflect the time-series characteristics of photovoltaic output and load fluctuation, a time-series weighting factor was introduced to correct the influence of voltage deviations at different time periods, thereby forming a time-series comprehensive sensitivity index that determined the priority of energy storage system installation locations. On this basis, a multi-objective optimization model was formulated that simultaneously minimized the comprehensive operating costs of the energy storage system, voltage deviation of the distribution network, and total network losses. Subsequently, the optimization model was solved using an improved particle swarm optimization algorithm, which enhanced convergence speed and avoided local optimal by dynamically adjusting inertia weight and learning factors. Finally, generative adversarial network was employed to generate diverse distribution network operational scenarios, thereby enriching the optimization dataset and improving the robustness of solutions under photovoltaic and load uncertainty. [Results] The case study results demonstrated that, compared with conventional voltage sensitivity analysis method, the proposed time-series sensitivity analysis method effectively enhanced the precision and targeted nature of node voltage regulation. The optimized energy storage system configuration achieved collaborative improvement in both economic performance and renewable energy utilization capability while ensuring that nodal voltages remained within permissible limits. [Conclusion] The proposed strategy significantly accelerates the energy storage planning process of distribution network, enhances its ability to maintain voltage stability under photovoltaic fluctuations, and promotes the economic and reliable operation of distribution network with high proportion photovoltaic.