Air target tracking by homing missile using deep learning and fuzzy adaptive control

In the integrated guidance and control approach, the guidance law and the autopilot are traditionally developed and tested separately, assuming the ideality of each other. This paper presents the design and simulation of a deep and fuzzy learning adaptive controller to guide a homing missile in a three-dimensional scenario to minimize collision time and maximize target interception accuracy. A deep learning neural network controller is initially developed offline in the proposed controller design and utilized as a gain table within the adaptive control framework. Subsequently, introducing fuzzy control further enhances the controller's adaptability and performance. The effectiveness of both controllers is evaluated under disturbance conditions. Simulation results demonstrate that the proposed controllers, along with the integrated guidance and control model, achieve reduced final miss distance and collision time compared to conventional PID and LQR controllers.