At present, the development of distributed photovoltaic in the entire county is progressing in an orderly manner. However, due tothe small and dispersed load volume of the distribution network in remote areas, the large-scale integration of distributed photovoltaic intothe distribution network brings uncertainty to the system operation, and is prone to risks of exceeding limits such as high or low voltage. Aprobabilistic evaluation method for system voltage and branch power flow to quantify the risk of exceeding limits is introduced. At the sametime, a two-layer optimization scheduling model for mobile energy storage in distribution networks considering the risk of exceeding limitswas established, and a dynamic stochastic optimal power flow algorithm was adopted to solve the established model. Taking the IEEE 33node system as an example, the scheduling control strategy of mobile energy storage is analyzed. The calculation results show that usingthe optimization scheduling method described in this article can not only ensure high operational efficiency of mobile energy storage, butalso improve the voltage quality qualification rate of the distribution system, reduce system network losses, and improve system power sup?ply reliability.