With the increase in the installed capacity of distributed energy resources and the continuous growth of electricity consumption for agricultural irrigation

safety issues such as line overload and voltage limit violations have emerged in rural power grids. To address these issues

an optimal scheduling method for rural power grids based on multi-agent compromise solutions is proposed to achieve collaborative optimal operation among multiple agents in rural power grids. First

three objective functions are established: maximizing the revenue of rural power grid operators

maximizing the revenue of distributed energy generators

and minimizing the electricity cost of irrigation loads. Then

a multi-agent optimal scheduling model is constructed for the three subjects. The multi-agent proximal policy optimization (MAPPO) algorithm and the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method are used to obtain the optimal compromise solution among multiple agents. During the optimization process

the MAPPO algorithm incorporates the action information of adjacent agents into the current state space to enhance collaboration among different agents

and the TOPSIS method is then used to determine the optimal scheduling scheme. The simulation results of a practical case study show that

compared with traditional algorithms

the proposed method improves the benefits of rural power grid operators and distributed energy generators

reduced the electricity cost of agricultural irrigation and processing time.