To address the problems of transformer overload caused by seasonal fluctuations in agricultural load and insufficient power distribution capacity in agricultural irrigation areas, a transformer siting method based on spatial coordinate system and genetic algorithm is proposed. First, agricultural irrigation data from a certain region in Jilin Province are screened and the spatial coordinate system is optimized. Then, with the goal of minimizing the weighted sum of social benefit cost, network loss cost, production cost of MMTS (Modular Mobile Transformer System) and travel cost, a siting model is constructed by combining the transformer loss model and travel cost model. MMTS is introduced to optimize transformer deployment for adapting to changes in power demand. The genetic algorithm is applied, with consideration of hourly local optimization, which reduces computational complexity and improves siting efficiency while sacrificing a small degree of accuracy. Case verification shows that when 4 MMTS units are configured, the objective function value is minimized, saving 110. 211 yuan in cost per day and achieving an annualized benefit of 39 240 yuan, which can significantly reduce

the load and loss of the service areas. After optimizing the spatial coordinate system, the hourly calculation time is only 1 083. 775 s, and the maximum error compared with the 24 h global calculation result is only 2. 175% , achieving a good balance between efficiency and accuracy.

site selection, modular mobile transformer system ( MMTS ), agricultural irrigation, genetic algorithm, optimization method of spatial coordinate system