Fuzzy Control of Multi-scale Target Tracking for Quadrotor Drones

Quadrotor drones face multiple challenges such as accuracy and real-time performance when tracking targets in a constantly changing and dynamic environment. To improve the target tracking accuracy and flight control stability of quadrotor drones in dynamic scenes, a quadrotor drone control method combining multi-scale target tracking algorithm and Type-2 fuzzy control is proposed. Firstly, a multi-scale object detection method based on kernel correlation filter is adopted, which can effectively cope with target scale and position changes through multi-scale analysis. Second, employing Type-2 fuzzy control to handle uncertainties in the control process ensures that the quadcopter drone accurately adjusts its flight state during target tracking. Experimental results show that the accuracy of the multi-scale object detection method based on kernel correlation filter is 0.97 on 1600 datasets. In terms of the comprehensive performance of target tracking and flight control, the accuracy of the Type-2 fuzzy control model is 0.92, the precision is 0.91, the recall rate is 0.91, the F1 value is 0.90, and the area under the curve value reaches 0.93, demonstrating strong target tracking ability and control accuracy. Experimental results show that the proposed multi-scale target tracking fuzzy control for quadrotor drones has excellent performance.