Adaptive Neural Network Control Research Articles

Identification and Control of Flexible Joint Robots Based on a Composite-Learning Optimal Bounded Ellipsoid Algorithm and Prescribe Performance Control Technique

This paper presents an indirect adaptive neural network (NN) control algorithm tailored for flexible joint robots (FJRs), aimed at achieving desired transient and steady-state performance. To simplify the controller design process, the original higher-order system is decomposed into two lower-order subsystems using the singular perturbation technique (SPT). NNs are then employed to reconstruct the aggregated uncertainties. An adaptive prescribed performance control (PPC) strategy and a continuous terminal sliding mode control strategy are introduced for the reduced slow subsystem and fast subsystem, respectively, to guarantee a specified convergence speed and steady-state accuracy for the closed-loop system. Additionally, a composite-learning optimal bounded ellipsoid algorithm (OBE)-based identification scheme is proposed to update the NN weights, where the tracking errors of the reduced slow and fast subsystems are integrated into the learning algorithm to enhance the identification and tracking performance. The stability of the closed-loop system is rigorously established using the Lyapunov approach. Simulations demonstrate the effectiveness of the proposed identification and control schemes.

Event-Triggered Adaptive Neural Network Control for State-Constrained Pure-Feedback Fractional-Order Nonlinear Systems with Input Delay and Saturation

In this research, the adaptive event-triggered neural network controller design problem is investigated for a class of state-constrained pure-feedback fractional-order nonlinear systems (FONSs) with external disturbances, unknown actuator saturation, and input delay. An auxiliary compensation function based on the integral function of the input signal is presented to handle input delay. The barrier Lyapunov function (BLF) is utilized to deal with state constraints, and the event-triggered strategy is applied to overcome the communication burden from the limited communication resources. By the utilization of a backstepping scheme and radial basis function neural network, an adaptive event-triggered neural state-feedback stabilization controller is constructed, in which the fractional-order dynamic surface filters are employed to reduce the computational burden from the recursive procedure. It is proven that with the fractional-order Lyapunov analysis, all the solutions of the closed-loop system are bounded, and the tracking error can converge to a small interval around the zero, while the state constraint is satisfied and the Zeno behavior can be strictly ruled out. Two examples are finally given to show the effectiveness of the proposed control strategy.

Adaptive neural network control of robotic manipulators with input constraints and without velocity measurements

AbstractThis paper addresses the trajectory tracking problem for a class of uncertain manipulator systems under the effect of external disturbances. The main challenges lie in the input constraints and the lack of measurements of joint velocities. An extend‐state‐observer is utilized to estimate the velocity signals; then, a neural‐network‐based adaptive controller is proposed to solve the problem, where a term based on the nominal model is included to enhance the tracking ability, and the effect of uncertainties and disturbances are compensated by a neural‐network term. Compared with the existing methods, the main distinctive features of the presented approach are: (i) The control law is guaranteed to be bounded by design, instead of directly bounded by a saturation function. (ii) The trade‐off between the performance and robustness of the presented controller can be easily tuned by a parameter that depends on the size of model uncertainties and external disturbances. By virtue of the Lyapunov theorem, the convergence properties of the proposed controller are rigorously proved. The performance of the controller is validated via both simulations and experiments conducted on a two‐degree‐of‐freedom robot manipulator.

Practical fixed-time neural control for MIMO non-strict feedback nonlinear systems: an adaptive neural network approach

ABSTRACT This paper studies the non-singular practical fixed-time neural adaptive control issues for multi-input and multi-output (MIMO) nonlinear systems with non-strict feedback form. Neural networks (NN) are used to estimate the unknown nonlinearities and deal with the problem of an algebraic loop. Under the framework of the backstepping control design, a practical fixed-time adaptive NN control method is developed by using the adding power integration technology. According to the Lyapunov function theory, it is proved that the closed-loop system is practical fixed-time stable, and the system can track the desired reference signal within a fixed time. Finally, the proposed practical fixed-time control method is applied to a multi-motor control platform, which proves the effectiveness of the control method.

Output-mask-based adaptive NN control for stochastic time-delayed multi-agent systems with a unified event-triggered approach

This paper investigates a class of output-mask-based adaptive neural network (NN) tracking control for nonlinear stochastic time-delayed multi-agent systems (STMASs) based on a unified event-triggered approach. The output signal relies on an output mapping acted as a mask, defined as a privacy-protection-like method, so that the internal state of one agent cannot be identified by other distrustful eavesdroppers or attackers. Moreover, the construction of a unified event-triggered control scheme retains the advantages of the saturation threshold triggering strategy, incorporates distributed errors, and increases the flexibility of thresholds. Furthermore, for stochastic time-delay multi-agent systems, the initial value limitation of the conventional first-order filter is removed by a first-order Levant differentiator, and a new estimation term in the fuzzy observer is established to solve the nonlinear fault. The unknown function in pure-feedback form is addressed via combining Butterworth low-pass filter and radial basis function neural networks (RBF NNs). Finally, the boundedness of all signals in the closed-loop systems is demonstrated, and the effectiveness of the proposed algorithm is verified by some simulation results.

A Continuous Neural Network Adaptive Controller for Consensus of Uncertain Multi-Agent Systems

A Continuous Neural Network Adaptive Controller for Consensus of Uncertain Multi-Agent Systems

Fuzzy neural network adaptive AUV control based on FTHGO

ABSTRACT A fuzzy radial basis function neural network (Fuzzy RBFNN) adaptive control scheme, based on fixed-time high gain state observer (FTHGO), is proposed to address the unpredictability of real-time state and composite interference in the trajectory tracking of the fully driven Autonomous Underwater Vehicle (AUV), ensuring fixed-time system convergence regardless of initial conditions. Firstly, a fixed-time backstepping controller is designed and a first-order fixed-time filter is introduced to tackle the differential explosion issue. Secondly, an FTHGO is developed to observe the real-time states of the AUV without assuming global known state signal. Then, the composite interference in the AUV system is effectively compensated by integrating the Fuzzy RBFNN technique. Finally, the fixed-time stability of the entire closed-loop system is proven utilising the Lyapunov stability theory. The effectiveness of the proposed algorithm is proved by the simulation experiment.

Resilient event-triggering adaptive neural network control for networked systems under mixed cyber attacks

This paper addresses the resilient event-triggering adaptive neural network (NN) control problem for networked control systems under mixed cyber attacks. Compared with the conventional event-triggered mechanism (ETM) with constant threshold, a novel resilient ETM is designed to withstand the affect of denial-of-service attacks and conserve communication resources. Different from the energy-bounded deception attacks, an unknown state-dependent nonlinear attack signal is considered in this work. To identify the deception attack, the NN technique is utilized to approximate the unknown attack signal. Subsequently, an adaptive controller is established to compensate for the malicious affects of deception attacks on the system. Furthermore, sufficient conditions for the boundedness of the system are derived via applying the Lyapunov functional, and a co-design strategy for control gain and event-triggering parameter is provided. Finally, the feasibility of the proposed approach is validated through a robot manipulator system.

Adaptive neural network controller for the rotating SMA actuator

In this research, we present a new data-driven Adaptive Neural Network (ANN) controller designed for a unique arc-shaped Shape Memory Alloy (SMA) actuator. The actuator generates rotational motion through two-dimensional shape restoration, which is significantly different from traditional linear SMA actuators. The SMA in an arc-shaped pattern tries to recover its original shape when heated, ultimately leading to rotational motion. Due to complexity in SMA motion, it is difficult to apply conventional empirical modeling methods in this research. Given this modelling difficulty, a model-based control approach may not be promising. Thus, a data-driven ANN is adopted here to control the SMA actuator without relying on precise modelling. This controller adeptly learns the actuator's dynamic behavior in real-time, fine-tuning its neural network weights to ensure optimal control. To validate its efficacy, we compare performance of the ANN controller with that of the traditional Proportional-Integral-Derivative (PID) and sliding mode controllers across two reference inputs (sinusoidal and square) under four different disturbance scenarios (input-output, output-only, input-only, and no disturbance). Our experimental results show the ANN controller can provide similar or slightly better performance in terms of tracking accuracy and disturbance rejection without rigorous parameter or gain tuning compared to traditional controllers. Our results verify advanced learning capabilities of the ANN controller and its potential in control of the SMA actuator.

Adaptive Neural Network Tracking Control of Robotic Manipulators Based on Disturbance Observer

This article presents an adaptive neural network (ANN) control scheme based on a disturbance observer that can achieve trajectory tracking control of robotic manipulators under external disturbances and dynamic model uncertainties. Firstly, an ANN controller based on full-state feedback is derived using the backstepping technique to achieve an online approximation of uncertainty. The integral sliding mode surface with a position error is introduced into the controller, which reduces the steady-state error of the system and enhances robustness. Then, a novel disturbance observer is designed to estimate both the approximation errors of the ANN and external disturbances, and to provide compensation for the controller, effectively suppressing the trajectory tracking errors caused by approximation errors and disturbances. Subsequently, the Lyapunov stability theory is utilized to demonstrate the stability of the developed control strategy and the boundedness of all closed-loop signals. Finally, numerical simulations are used to confirm the efficacy of the proposed control method.

Model-free based adaptive BackStepping-Super Twisting-RBF neural network control with α-variable for 10 DOF lower limb exoskeleton

Model-free based adaptive BackStepping-Super Twisting-RBF neural network control with α-variable for 10 DOF lower limb exoskeleton

Neural network‐based output‐feedback adaptive control for a class of uncertain strict feedback fractional‐order nonlinear systems subject to input saturation

SummaryThis paper presents neural networks (NNs) adaptive controller for an uncertain fractional‐order nonlinear system in strict‐feedback form, subject to input saturation, unavailable states for measurement, and external disturbances. The fractional‐order adaptive laws are derived based only on the output tracking error thanks to the implementation of the strictly positive real (SPR) property, differently from the existing results in the literature of fractional‐order strict‐feedback systems where all system states are used in the adaptive laws. The proposed design approach addresses the nonaffine nature of the control input due to saturation nonlinearity by using the mean value theorem and follows a nonrecursive design by using a state transformation where the advantage is twofold. First, it eliminates the explosion complexity found in back‐stepping control‐based approaches design, and second, it reduces the approximators units and parameters in controller implementation. Furthermore, an observer is introduced to estimate the unavailable newly defined states, and then an output adaptive feedback control design is ensured. An NN is used to approximate the unknown ideal control law, and an auxiliary control term is appended to deal with saturation effect, unknown disturbance, and approximation errors. The tracking error is proved to converge asymptotically to a bounded set using the Lyapunov theory. Simulation results on three examples show the effectiveness of the proposed approach.

Adaptive neural network anti-disturbance control of a cable-driven flexible-joint-based underwater vehicle manipulator system with dead-zone input nonlinearities via multiple neural observers

To achieve the requirements of lightweight, low energy consumption, and low inertia of an underwater vehicle manipulator system, a cable-driven manipulator is installed on the underwater vehicle to form a cable-driven flexible-joint-based underwater vehicle manipulator system (CDFJ–UVMS). The CDFJ–UVMS is a complex nonlinear system subject to model uncertainties, complex marine environment disturbances, and actuator dead-zone nonlinearity. To design track controllers, the CDFJ–UVMS dynamics is divided into two parts: known and unknown. Subsequently, a radial basis function neural network is adopted to approximate the unknown nonlinearity. A neural network performance observer is constructed, whose estimation error is then used to design a novel neural disturbance observer (NDO) to estimate the total disturbance. Finally, an adaptive neural network control method is proposed for the CDFJ–UVMS based on the NDO, neural network compensator, and neural performance observer. The stability of the closed-loop system is analyzed using the Lyapunov method. The proposed control algorithm is applied to a CDFJ–UVMS with two cable-driven joints and compared with other control methods to show the effectiveness of the proposed control algorithm.

Neural network adaptive control for turbo‐generator of power systems with prescribed performance and unknown asymmetric actuator dead zone

AbstractThis paper investigates a neural network adaptive controller design method for turbo‐generator of power systems with external disturbance, unknown system dynamic, prescribed performance, and unknown actuator dead zone. First, the unknown system dynamic is estimated and overcome through a neural network. Using the implicit function theorem, the unknown asymmetric dead‐zone behavior of the actuator is overcome by another static neural network. Second, the external disturbance and the reconstruction error of neural networks are handled by a robust term updated online. Moreover, there is no requirement to pre‐know or off‐line estimate the reconstruction error of neural networks and the upper bound of the system external disturbance. Third, based on Lyapunov theory, the smooth control law is proposed. In the meantime, the uniform ultimate boundedness of the networks weights is strictly proved. Furthermore, the system state error satisfies the prescribed transient performance and can converge to a small neighborhood around zero. Finally, a numerical simulation shows the effectiveness of the proposed method.

Decentralized event‐triggered output feedback adaptive neural network control for a class of MIMO uncertain strict‐feedback nonlinear systems with input saturation

SummaryFor a class of multiple‐input multiple‐output large‐scale nonlinear systems in strict‐feedback form with input saturation, external disturbances and immeasurable states, an adaptive decentralized neural network (NN) control strategy on the basis of event triggered mechanism is investigated in this article. In contrast to the literature, the proposed method is centered on the control‐error as a replacement to the tracking‐error that leads to a simplified derivation approach of adaptive laws. Furthermore, the control gains for this class of systems are considered unknown nonlinear functions and not assumed as simple unity or known gains as always done in the literature. Moreover, the challenge of losing controllability that typically arises in state transformation‐based methods, as reported in the literature, is entirely resolved in our approach. Last and not least, all restrictions imposed on unmatched interconnections are eliminated along with avoiding the complexity explosion caused by recursive back‐stepping designs. For this end, the unknown ideal control laws are approximated using NNs, while additional control terms are added to handle saturation effects, unknown interactions, and approximation errors. Additionally, fuzzy inference systems are employed to estimate unknown control errors. Due to the strictly positive real property, the tracking errors are proved to belong to a small compact set using Lyapunov theory. Simulation results demonstrate the effectiveness of the proposed approach.

Adaptive self-structuring neural network control for full-state constrained hydraulic systems with disturbance compensation

An adaptive self-structuring neural network for full-state constrained systems with a prescribed performance control scheme is proposed for hydraulic systems to achieve asymptotic tracking under parametric uncertainties and time-varying disturbances. To overcome external disturbances in hydraulic systems, a self-structuring neural network is employed to approximate mismatched disturbances, reducing the computational burden. Additionally, an output feedback filtering control technique is introduced to address the challenges of uncertain parameter sets, which deteriorates the control performance when affected by oscillations triggered by high-frequency noise. The prescribed performance control achieves the desired transient and steady-state performance of the system by constructing a sequence of error transition variables that guarantee that all the state errors of the hydraulic systems are within the appointed-time reachable performance function boundary range. For the proposed control scheme, the stability of the closed-loop system can be demonstrated via Lyapunov theory, and all signals are guaranteed to be bounded. Ultimately, the tracking performance of the proposed controller is verified through abundant comparative experiments under the influence of parameter uncertainty and time-varying disturbances.

Maximum Power Point Tracking Control of Offshore Hydraulic Wind Turbine Based on Radial Basis Function Neural Network

A maximum power point tracking control strategy for an affine nonlinear constant displacement pump-variable hydraulic motor actuation system with parameter uncertainty, used within an offshore hydraulic wind turbine, is studied in this paper. First, we used the feedback linearization method to solve the affine nonlinear problem in the system. However, offshore hydraulic wind turbines have strong parameter uncertainty characteristics. This conflict was resolved through the further application of RBF neural network adaptive control theory. So, we combined feedback linearization with RBF adaptive control as the control theory, and then two control laws were compared by setting the pump rate and rating as outputs, respectively. It is shown by the MATLABR2016a/Simulink emulation results that power control is smoother than speed and friendlier for electric networks. It is also shown by the emulation results, in terms of the undulatory wind speed condition, that the feedback linearization–RBF neural network adaptive control strategy has perfect robustness. According to the simulation results, the feedback linearization–RBF neural network adaptive control strategy adopts the RBF neural network to approach complex nonlinear models and solve the parameter uncertainty problem. This control law also avoids the use of feedback linearization control alone, which can result in the system becoming out of control.

Robot skill learning system of multi-space fusion based on dynamic movement primitives and adaptive neural network control

This article develops a robot skill learning system with multi-space fusion, simultaneously considering motion/stiffness generation and trajectory tracking. To begin with, surface electromyography (sEMG) signals from the human arm is captured based on the MYO armband to estimate endpoint stiffness. Gaussian Process Regression (GPR) is combined with dynamic movement primitive (DMP) to extract more skills features from multi-demonstrations. Then, the traditional DMP formulation is improved based on the Riemannian metric to encode the robot’s quaternions with non-Euclidean properties. Furthermore, an adaptive neural network (NN)-based finite-time admittance controller is designed to track the trajectory generated by the motion model and to reflect the learned stiffness characteristics. In this controller, a radial basis function neural network (RBFNN) is employed to compensate for the uncertainty of the robot dynamics. Finally, experimental validation is conducted using the ROKAE collaborative robot, confirming the effectiveness of the proposed approach. In summary, the presented framework is suitable for human-robot skill transfer method that require simultaneous consideration of position and stiffness in Euclidean space, as well as orientation on Riemannian manifolds.

FTESO-adaptive neural network based safety control for a quadrotor UAV under multiple disturbances: algorithm and experiments

PurposeQuadrotors have been applied in various fields. However, because the quadrotor is subject to multiple disturbances, consisting of external disturbances, actuator faults and parameter uncertainties, it is difficult to control the unmanned aerial vehicle (UAV) to achieve high-precision tracking performance. This paper aims to design a safety controller that uses observer and neural network method to improve the tracking performance of UAV under multiple disturbances. The experiments prove that this method is effective.Design/methodology/approachFirst, to actively estimate and compensate the synthetic uncertainties of the system, a finite-time extended state observer is investigated, and the disturbances are transformed into the extended state of the system for estimation. Second, an adaptive neural network controller that does not accurately require the dynamic model knowledge is designed based on the estimated value, where the weights of the neural network can be dynamically adjusted by the adaptive law. Furthermore, the finite-time bounded convergence of the proposed observer and the stability of the system are proved through homogeneous theory and Lyapunov method.FindingsThe figure-“8” climbing flight simulation and real flight experiments illustrate that the proposed safety control strategy has good tracking performance.Originality/valueThis paper proposes the safety control structure of the UAV, which combines the extended state observer with the neural network method. Numerical simulation results and actual flight experiments demonstrate the effectiveness of the proposed control strategy.

Neural networks-based adaptive command filter control for nonlinear systems with unknown backlash-like hysteresis and its application to single link robot manipulator

<abstract><p>In this paper, an adaptive neural network control problem for nonstrict-feedback nonlinear systems with an unknown backlash-like hysteresis and bounded disturbance was presented. Radial basis function neural networks (RBFNN) were used to approximate the unknown functions and the problem of the explosion of complexity problem was handled by utilizing the command filter method. Furthermore, the influence of an unknown backlash-like hysteresis input was addressed by approximating an intermediate variable. Based on the backstepping method and the command filter technique, an adaptive neural network controller was designed via the approximation abilities of RBFNN. With the help of the Lyapunov stability theory, the proposed controller ensures that all of the signals in closed-loop systems are bounded and that the tracking error fluctuates close to the origin within a bounded area. Finally, a real-world example based on the single-link manipulator was shown to demonstrate the viability of the presented approach.</p></abstract>