A two-stage mixed integer programming model was established to solve the problem of fine match of task allocation and scheduling in the “single base station multiple drones” mode of plant protection drones on the platform of the agricultural socialization service. First， the cost minimization goal of task allocation and scheduling of plant protection drone was solved under the condition of quickly and accurately solving the dispersed plant protection needs of farmers by introducing agricultural parameter variables including variable time windows for pesticide efficacy under the influence of temperature， flight mode selection of crop protection drone in farmland， and variable power consumption rate. Then， the improvement of the cost minimization goal was studied by changing the rated battery life time， flight mode in the field， optimal time window of drug efficacy corresponding to temperature and the size of demand of the plant protection drone while ensuring other parameters remain unchanged. The results showed that plant protection drones reduced the plant protection cost by 40.02%，12.45%，21.17% and 39.19%，respectively， by carrying batteries with long duration of flight，choosing the long-side turning grid flight， selecting pesticide types with the best efficacy and a wide range of temperature， and increasing the upper limit of platform service volume per unit time period. It is indicated that the mixed Integer programming model， which introduces specific agricultural production factors， is scientific and effective， and is more suitable for the task allocation and scheduling of plant protection drones on the platform of agricultural service.