–Brief Paper– NONLINEAR AND ADAPTIVE INTELLIGENT CONTROL TECHNIQUES FOR QUADROTOR UAV – A SURVEY Hongwei Mo and Ghulam Farid ABSTRACT Parametric uncertainties and coupled nonlinear dynamics are inherent in quadrotor configuration and infer adaptive nonlinear ap- proaches to be used for flight control system. Numerous adaptive nonlinear and intelligent control techniques, which have been reported in the literature for designing quadrotor flight controller by various researchers, are investigated in this paper. As a priori, each conventional nonlinear control technique is discussed broadly and then its adaptive/observer based augmentation is conferred along with all possible var- iants. Among conventional nonlinear control approaches, feedback linearization, backstepping, sliding mode, and model predictive control, are studied. Intelligent control approaches incorporating fuzzy logic and neural networks are also discussed. In addition to adaption based parametric uncertainty handling, various other aspects of each control technique regarding stability, disturbance rejection, response time, asymptotic, exponential and finite time convergence etc., are discussed in sufficient depth. The contribution of this paper is the provision of detailed and in depth discussion on quadrotor nonlinear control approaches to the flight control designers. Key Words: UAV, Quadrotor, Nonlinear control, Intelligent control, Flight controller. I. INTRODUCTION Availability of low-cost inertial measurement units (IMUs) has made it possible to design a variety of miniature unmanned aerial vehicles (UAVs) [1]. Among them, the utmost nascent and intriguing UAV is the quadrotor, which has the ability of vertical takeoff and landing (VTOL). The hovering ability of quadrotor ae- rial crafts makes them superior to fixed-wing UAVs. The latter are inept at hovering and cannot be engaged in environments where stationary or quasi-stationary flights are required [2]. Quadrotor is widely used, both for military and civil purposes. Therefore, a lot of funding and research interests have been shown by different communities to take it a step further. Several projects on quadrotor UAV by various researchers havebeen accomplished since the last decade. Quadrotor has recently been used for surveillance, aerial photography, building exploration, climate forecasting, bridge in- spection, 3D mapping and swarm missions etc. [2–5]. The quadrotor UAV holds very simple mechanical structure, while on the contrary, conventional helicopter normally retains variable pitch rotor and extremely complex mechanical control structure. Quadrotor has cross configuration and there are four fixed-pitch rotors at each end. Referring to Fig. 1, rotors 1 and 3 constitute an odd pair while rotors 2 and 4 constitute an even pair. Both pairs of rotors rotate contrarily to balance the total rotating moment of the body. The output states to be controlled are more than available inputs in a quadrotor, therefore, it constitutes an underactuated system (fewer actuators having more DOF to be controlled) and embraces inherent instability. Six output states which need to be controlled are three Cartesian position states (x, y , and z) and three Euler angles (roll, pitch, and yaw). Available inputs are thrust and torque terms generated by four different speed combinations of propellers. Numerous commercially available and open source quadrotors are classical PID controller based and exist in different variants [6]. These controllers are well suited for stationary or quasi-stationary flights where the dynamics act as a linearized model [7]. Linearization of the model puts constraints on the operating re- gime and these controllers work only near the hovering state. In most of the practical applications system parameters are well known, and can be used to determine PID controller parame- ters but still there present some unknown parts e.g. wind gusts, and sensor spoofing etc. Therefore, it may become challenging, for such controllers that purely rely on precise plant dynamics, to show good path following performance under the existence of unknown perturbations and wind turbulence. In addition to possible uncer- tain parameters and varying wind gusts, quadrotor owns underactuated nonlinear dynamics. Aggressive quadrotor maneu- vers exhibit a strong nonlinear behavior, therefore, linear control- lers fail to converge and nonlinear solutions turn out to be inevitable. Moreover, parametric uncertainties and unmodeled non- linear dynamics are innate in quadrotor structure which arise due to variation in mass and mass moments of inertia. This calls for adap- tive and intelligent nonlinear control approaches for more precise control of quadrotor. In addition, quadrotor actuator uncertainties also require proper compensation. Adaptive rules have been inte- grated with various conventional nonlinear control approaches to make them more robust against mentioned uncertainties. All these adaption-based nonlinear control approaches are reviewed in this survey and a thorough discussion is presented. For prior in depth insight, this paper also discusses the conventional nonlinear control techniques for quadrotor flight control design. This survey study is intended to provide a profound insight, to the designers working on quadrotors, regarding various adaption-based intelligent and nonlinear control approaches. Moreover, the goal of this study is to analyze the existing flight Manuscript received April 12, 2017; revised October 29, 2017; accepted November 12, 2017. The authors are with College of Automation, Harbin Engineering University, China Ghulam Farid is the corresponding author (e-mail: farid.anjum@yahoo.com). Asian Journal of Control, Vol. 21, No. 3, pp. 1–20, March 2019 Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/asjc.1758 © 2018 Chinese Automatic Control Society and John Wiley & Sons Australia, Ltd