Research on RBF-PID hybrid drive control models for drone pesticide applications in tropical orchards

Durian is an important economic crop in Southeast Asia; it is susceptible to various pests and diseases, which affects its fruit quality and yield. Traditional manual and mechanical spraying methods have many shortcomings in spraying durian blind spots (such as tall and dense durian canopies). Drones can fly over these blind spots and carry out all-around pesticide spraying. How to combine artificial intelligence and automatic control to realize the spraying of pesticides by drones in the blind spots of durian trees has become a key research issue. Therefore, this study proposes an IM-PID (Improved Proportional-Integral-Derivative) control algorithm to improve the accuracy of drone pesticide spraying in the blind areas of durian tree pests and diseases. It introduces the RBF (Radial Basis Function) neural network to adjust the proportion, integral, and differential of incremental PID controller, which adjusts spraying parameters in real time to improve the accuracy of pesticide spraying. The experimental results show that the IM-PID control algorithm is superior to the traditional PID, fuzzy logic and sliding mode control algorithms regarding the spray flow accuracy, droplet distribution uniformity, and dynamic adjustment capabilities. It can significantly improve the efficiency of pesticide spraying in durian orchards and solve the problem of traditional spraying methods in blind areas, which controls pests and diseases and ensures the high quality and yield of durian fruits. It also reduces the environmental impact of excessive pesticide spraying and improves the economic benefits of durian cultivation.