Multi-access edge computing (MEC) uses unmanned aerial vehicles (UAVs) as edge nodes for dynamic deployment, which can cover complex environments and greatly enhance the application scope. Rotary-wing UVAs are now commonly used, but they cannot load heavy MEC servers. Therefore, the paper proposes the research of MEC of UAVs cooperation applied in power system. By taking advantage of the flexibility and high load of the fixed-wing UAV, it can be deployed in the environment with wide range, sparse distribution of power equipment(PE) and difficult communication, so as to support the use of PE. This is because the MEC servers loaded by UVAs have a low data processing capability. Therefore, tasks uploaded by PE often need to wait in a queue for a long time. However, the priorities of power services vary and cannot be processed equally. So combining real-time scheduling algorithm, simulated annealing(SA) algorithm and differential evolution(DE) algorithm, the paper proposes an MEC offloading algorithm assisted by UVAs for PE which have different priorities. By flexibly adjusting the order of power tasks in the waiting queue, the best result can be obtained under the conditions of meeting the delay requirements of all PE and making the high priority tasks process faster.