ABSTRACT
After modern control theories have been widely applied in aviation and aerospace field, the scholars worldwide, aiming at solving the problems of SISO
design issues, launch studies of using modern control theory to the design of BTT missile autopilot. Lin etc. designed a general autopilot for a BankTo-Turn (BTT) missile, which is through a generalized Linear Quadratic Gaussian and Loop Transfer Recovery (LQG/LTR) method. The detailed design procedure including two steps, Kalman filter design and an optimal control law design. The proposed controllers have a prescribed degree of stability even in the case of nonminimum phase problem[5]. Meanwhile, some scholars use feedback linearization method to linearize the nonlinear model of BTT missile and design the control system according to linear system theory. Feedback linearization is one of the important approaches of designing the nonlinear control system, dividing into nonlinear dynamic inversion method and differential geometry method[6]. In such two methods, nonlinear dynamic inversion is not limited to the specific form of system equations, which makes it available to investigate general nonlinear system directly. Also its physical concept is clear and intuitive, so this method is widely employed in nonlinear control engineering[7,8]. However, this method cannot deeply analyze the inherent characteristics of nonlinear systems. With the increase of complexity of nonlinear system, the uncertainty of nonlinear dynamic inverse system raise. Therefore in recent years, some
scholars have used the differential geometry feedback linearization method to design BTT missile control system. Zhang etc. designed a BTT missile controller based on differential geometry feedback linearization method. Taking the existence of internal dynamics subsystem in the full-state dynamics model into consideration, the BTT missile dynamics model is simplified through the ignorance of lift and lateral forces generated by deflection of rudder surfaces. Also the exact linearization system whose relative order is equal to the dimension of the system is achieved. Then with the linear state feedback, the command orders which can steadily control roll angle, attack angle and side-slip angle are given[9].
