To research the relay selection of wireless sensor network ( WSN) to improve the monitoring level of distribution network

a self-adaptive ε-greedy algorithm-based WSN relay selection optimization method is proposed according to the existing ε- greedy method. Firstly

considering the complex topology of distribution network and the application scenarios of strong electromagnetic interference

this paper modelled a WSN communication system model. Secondly

the dynamic relay selection problem of WSN supporting distribution network monitoring is constructed. Then

adopting the accumulated reward value of the historical strategies to adaptively adjust the exploration intensity to minimize network energy consumption and ensure reliable data transmission. Finally

the simulation results show that compared with the descending ε-greedy algorithm

the traditional ε-greedy algorithm and the shortest path method

the proposed algorithm can reduce energy consumption by 8. 23%

12. 85% and 17. 11%

respectively.