A large number of distributed generators (DG), such as wind power generators and photovoltaic cells, will be connected to the power distribution network, resulting in voltage fluctuation and active network loss increase. The dynamic reactive power needs to be optimized. However, wind-solar uncertainty will affect the effect of dynamic reactive power optimization. Thus this paper proposes a multi-objective intelligent optimization method for the dynamic reactive power of a distribution network with a solid state transformer (SST) considering wind-solar uncertainty. First, the wind speed and light intensity are fitted by the Weibull and Beta distributions, and then the output formulas of wind power generators and photovoltaic cells are used to generate the DG output model. Second, Monte Carlo simulation sampling is used to sample the above model to generate thousands of daily DG output scenarios, and the k-means clustering algorithm is used to cluster thousands of scenarios into several typical scenarios to shorten the calculation time of random power flow. Third, based on the IEEE33 node system, an active distribution network scheme with the SST and an active distribution network scheme with on-load tap changer transformer are established. To minimize the active power loss and node voltage fluctuation of the distribution network, the relevant parameters of the two schemes are optimized using an improved multi-objective gray wolf optimizer (MOGWO) algorithm. Finally, the optimized parameters are simulated and compared to prove the superiority of the proposed method in reducing network loss and maintaining node voltage stability.