Tracking control problem of nonlinear strict-feedback systems with input nonlinearity: An adaptive neural network dynamic surface control method.

A novel adaptive neural network dynamic surface control is proposed for hypersonic vehicle. Based on the hypersonic vehicle model characteristics, the adaptive neural network dynamic surface control attitude controller and the neural networks velocity controller are designed, respectively. By combining Nussbaum gain function with decoupled backstepping, the problems of unknown control directions and control singularity from hypersonic vehicles model in strict-feedback form are simultaneously solved. Furthermore, the norm of the ideal weighting vector in neural network systems is considered as the estimation parameter, such that only one parameter is adjusted at each recursive step. Based on decoupled backstepping method and Lyapunov stability theorem, the semi-global stability of the flight control system is proved. Simulation results show that the new controller can guarantee the expected tracking performance.

In this paper, a longitudinal model of hypersonic vehicle with variable geometry inlet is established. Because of the strict requirements for the angle of attack of the scramjet during the flight of the aircraft, the uncertainty introduced by the parameter fitting, the rotating lip cover in the longitudinal model, and the uncertain external interference of the aircraft. The dynamic surface control technology is used to design the angle of attack autopilot of the aircraft and the Radial basis function (RBF) neural network is used to realize the adaptive approximation of the uncertain part of the model, to suppress the interference and accurately track the instructions. Finally, the simulation results show that the method can effectively control the angle of attack of hypersonic vehicle with variable geometry inlet, meet the performance requirements and verify the correctness of the method.

This paper focuses on the problem of adaptive output-constrained neural tracking control for uncertain nonstrict-feedback systems in the presence of unknown symmetric output dead zone and input saturation. A Nussbaum-type function-based dead-zone model is introduced such that the dynamic surface control approach can be used for controller design. The variable separation technique is employed to decompose the unknown function of entire states in each subsystem into a series of smooth functions. Radial basis function neural networks are utilized to approximate the unknown black-box functions derived from Young's inequality. With the help of auxiliary first-order filters, the dimensions of neural network input are reduced in each recursive design. A main advantage of the proposed method is that for an -order nonlinear system, only one adaptation parameter needs to be tuned online. It is rigorously shown that the proposed output-constrained controller guarantees that all the closed-loop signals are semiglobal uniformly ultimately bounded and the tracking error never violates the output constraint.

Pneumatic servo system is widely applied in many industries, which has advantages comparing with electromechanical and hydraulic system because of its fast response, high-performance quality and low-cost. However, the servo control methods for pneumatic system still have some inevitable drawbacks and problems remaining to be researched. In this paper, a position feedback dynamic surface control is designed which is based on our pneumatic actuator model. More importantly, in order to overcome model uncertainties, noise interference and external force disturbance, an adaptive neural network dynamic surface controller is proposed to overcome the negative effects. Besides, the stability of the pneumatic system is substantiated by Lyapunov stability theorem. Finally, the results of simulation experiment also prove that the adaptive neural network dynamic surface controller has more advantages than the traditional controllers in pneumatic position servo control.

This paper presents an adaptive neural tracking control approach for uncertain stochastic nonlinear time‐delay systems with input and output constraints. Firstly, the dynamic surface control (DSC) technique is incorporated into adaptive neural control framework to overcome the problem of ‘explosion of complexity’ in the control design. By employing a continuous differentiable asymmetric saturation model, the input constraint problem is solved. Secondly, the appropriate Lyapunov‐Krasovskii functional and the property of hyperbolic tangent functions are used to deal with the unknown time‐delay terms, RBF neural network is utilized to identify the unknown systems functions, and barrier Lyapunov functions (BLFs) are designed to avoid the violation of the output constraint. Finally, based on adaptive backstepping technique, an adaptive neural control method is proposed, and it decreases the number of learning parameters. Using Lyapunov stability theory, it is proved that the designed controller can ensure that all the signals in the closed‐loop system are 4‐Moment (or 2 Moment) semi‐globally uniformly ultimately bounded (SGUUB) and the tracking error converges to a small neighborhood of the origin. Two simulation examples are provided to further illustrate the effectiveness of the proposed approach.

An improved adaptive neural dynamic surface control for pure-feedback systems with full state constraints and disturbance

Fixed-time adaptive neural tracking control for high-order nonlinear switched systems with input saturation and dead-zone

In this paper, finite time adaptive neural network (NN) tracking control is investigated for a class of uncertain nonstrict feedback systems with unknown dead-zone and unmodeled dynamics as well as full state restrictions. Using the linearized representation of dead-zone, finite-time adaptive control is presented by the aid of dynamic surface control (DSC) technique and the property of Gaussian function. By introducing logarithmic function as invertible nonlinear mapping, the full state restriction problem is solved. The dynamic signal generated by the low-pass filter is used to handle unmodeled dynamics. Through finite-time stability theory and introducing the bounded closed set in the proof, all the signals in the closed-loop system are proved to be semi-globally practical finite time stable (SGPFS). The restriction conditions are not triggered for all the states. Finally, two numerical examples are used to illustrate the feasiability of the finite time adaptive control strategy.

A robust adaptive neural dynamic surface control (DSC) approach is presented for the longitudinal dynamics of a morphing aircraft in the presence of unknown dynamics and input constraints. For the altitude subsystem, neural systems are utilized to approximate the unknown nonlinear functions with smooth robust compensations to counteract the lumped approximation errors. By combining dynamic surface control and minimal learning parameter techniques, a robust adaptive neural control scheme is proposed and a simple adaptive algorithm is constructed. Meanwhile, an auxiliary system is incorporated into the control scheme to overcome the problem of input saturation. The highlight is that the proposed neural controller not only owns less updated neural parameters, but also has the ability of handling input constraints. It is proved that all the signals in the closed-loop system are bounded. Simulation results demonstrate the effectiveness of the proposed control scheme.

A robust adaptive tracking control problem is discussed for a class of strict-feedback uncertain nonlinear systems. Takagi-Sugeno type fuzzy logic systems are used to approximate the uncertainties. A unified and systematic procedure is developed to derive a novel robust adaptive tracking controller by use of the input-to-state stability (ISS) and by combining the dynamic surface control (DSC)-based backstepping technique and generalized small gain approach. The key features of the algorithm are that, firstly, the problem of ldquoexplosion of complexityrdquo inherent in the conventional backstepping method is circumvented, secondly, the number of parameters updated on line for each subsystem is reduced dramatically to 2. These features result in a much simpler algorithm, which is convenient to realize in application. In addition, it is shown that all closed-loop signals are semi-global uniformly ultimately bounded (SGUUB). Finally, simulation results via an application example of a pendulum system with motor is used to demonstrate the effectiveness and performance of the proposed scheme.

This paper is concerned with robust stabilization problem for a class of nonaffine pure-feedback systems with unknown time-delay functions and perturbed uncertainties. Novel continuous packaged functions are introduced in advance to remove unknown nonlinear terms deduced from perturbed uncertainties and unknown time-delay functions, which avoids the functions with control law to be approximated by radial basis function (RBF) neural networks. This technique combining implicit function and mean value theorems overcomes the difficulty in controlling the nonaffine pure-feedback systems. Dynamic surface control (DSC) is used to avoid "the explosion of complexity" in the backstepping design. Design difficulties from unknown time-delay functions are overcome using the function separation technique, the Lyapunov-Krasovskii functionals, and the desirable property of hyperbolic tangent functions. RBF neural networks are employed to approximate desired virtual controls and desired practical control. Under the proposed adaptive neural DSC, the number of adaptive parameters required is reduced significantly, and semiglobal uniform ultimate boundedness of all of the signals in the closed-loop system is guaranteed. Simulation studies are given to demonstrate the effectiveness of the proposed design scheme.

This paper introduces a novel control strategy for managing the uncertainties in flexible joint manipulators, incorporating a Radial Basis Function Neural Network (RBFNN) with Adaptive Dynamic Surface Control (ADSC). This strategy innovatively utilizes RBFNN to precisely approximate uncertain system dynamics and integrates a nonlinear damping term to effectively counteract external disturbances, enhancing the overall control accuracy. We have also developed an adaptive law that updates neural network weights and system parameters in real-time, ensuring the system's adaptability to dynamic changes. The application of the Lyapunov method ensures that all signals within the closed-loop system remain semi-globally uniformly bounded, significantly reducing tracking errors. Moreover, we introduce the use of Long Short-Term Memory (LSTM) networks for predictive analysis of state data, which further confirms the robustness and effectiveness of our control method through extensive simulations. The distinctive integration of these technologies and their practical validation through comparative simulations underscore the innovative aspects of our approach in addressing real-world challenges in flexible manipulators.

This paper presents a neural adaptive controller for a class of nonlinear system in the presence of input saturation and arbitrary unknown information by integrating the design framework of neural networks (NNs) with the dynamic surface control (DSC) and backstepping technique. NNs are directly used to deal with unknown nonlinear terms. By introducing a well-defined smooth function and a Nussbaum function, a novel robust adaptive control scheme is developed to guarantee the states of system to be semi-globally uniformly ultimately bounded (SGUUB). Owing to the introduction of DSC, the problem of ‘explosion of complexity’ inherent in the backstepping design is effectively eliminated. Simulation results demonstrate the effectiveness and simplicity of the proposed method.

This article focuses on an adaptive dynamic surface tracking control issue of nonlinear multiagent systems (MASs) with unmodeled dynamics and input quantization under predefined accuracy. Radial basis function neural networks (RBFNNs) are employed to estimate unknown nonlinear items. A dynamic signal is established to handle the trouble introduced by the unmodeled dynamics. Moreover, the predefined precision control is realized with the aid of two key functions. Unlike the existing works on nonlinear MASs with unmodeled dynamics, to avoid the issue of "explosion of complexity," the dynamic surface control (DSC) method is applied with the nonlinear filter. By using the designed controller, the consensus errors can gather to a precision assigned a priori. Finally, the simulation results are given to demonstrate the effectiveness of the proposed strategy.

This paper addresses the adaptive neural network tracking control problem for a class of strict-feedback systems with unknown non-linearly parameterised and time-varying disturbed function of known periods. Radial basis function neural network and Fourier series expansion are combined into a new function approximator to model each suitable disturbed function in systems. Dynamic surface control approach is used to solve the problem of ‘explosion of complexity’ in backstepping design procedure. The uniform boundedness of all closed-loop signals is guaranteed. The tracking error is proved to converge to a small residual set around the origin. A simulation example is provided to illustrate the effectiveness of the control scheme designed.