With the increase of renewable energy penetration, the power fluctuation of optical storage charging microgrids poses a serious challenge to the stability of distribution networks. In this paper, a comprehensive assessment framework based on computational analysis is proposed to quantify the static and dynamic impacts of new energy outputs on the distribution network by constructing a multi-microgrid system topology model and a windoptical-hydrogen cooperative regulation mechanism. The system-level and device-level constraint models are established by combining the connectivity graph theory, covering the branch voltage balance, node power limit, and the dynamic boundaries of the state of charge (SOC) of energy storage. A combined direct-indirect prediction method is designed to realize short-term and long-term prediction for the strong stochasticity of photovoltaic, wind turbine output and load power. The hydrogen production-storage-generation system is further introduced to smooth out new energy fluctuations by modeling hydrogen production and power regulation capability. Simulation experiments based on Matlab/Simulink show that the prediction accuracy of the proposed prediction model reaches 95.18% for photovoltaic output and 80.71% for wind turbine output. Under the optimized configuration strategy, the critical value of energy storage SOC is controlled at 89.45%, while the MPPT is 90.13%, which is more than 90% and is harmful to the safety of the energy storage device. The peak system load power is 606.41W, which is not exceeded. Comparing with KPCA, attention mechanism and other methods, the average assessment accuracy of voltage stability based on computational analysis is 99.18%, TSI=1.0, Gmean=99.48, which significantly improves the disturbance resistance of distribution network.
