Aiming at the problem of large load shedding caused by insufficient power supply capacity of rural distribution stations due to the superposition of photovoltaic output fluctuation and centralized operation of seasonal agricultural loads in rural distribution networks with high photovoltaic penetration, this paper proposes an optimal scheduling strategy for Interconnected Energy Storage Vehicles (IESV) oriented to power supply risk control in agricultural distribution networks. The system architecture of a rural distribution network based on IESV is constructed, and the mathematical model of IESV is established. The significant spatio-temporal fluctuation characteristics of agricultural loads are described, and on this basis, a two-stage scheduling strategy including pre-layout and dynamic dispatch is proposed. In the pre-layout stage, based on typical scenarios before the busy farming period, with the objectives of minimizing the configuration cost and load rate deviation penalty, the interconnected distribution station pairs are determined to reserve operation margin for the load peak during the busy farming period. On this basis, the dynamic dispatch stage is proposed to further cope with load fluctuations. Based on the topology formed by pre-layout, this stage optimizes the access locations and discharge power of IESV with the objective of minimizing the load shedding power. The two stages cooperate under the sequential optimization framework of "planning-operation" to realize the comprehensive optimization of operation cost and load shedding. Finally, the simulation based on the IEEE 33-node system verifies that the proposed strategy can significantly reduce the agricultural load shedding and the total system operation cost while controlling the power supply risk of agricultural distribution networks.