Adaptive Neural Network Control Research Articles

Quantized Input Robust Adaptive Neural Network Control for Nonlinear Systems With Full State Constraints

Quantized Input Robust Adaptive Neural Network Control for Nonlinear Systems With Full State Constraints

Finite-Time Consensus Adaptive Neural Network Control for Nonlinear Multiagent Systems Under PDE Models.

In this article, a novel adaptive control method based on neural networks is proposed for a class of multiagent systems (MASs) with nonlinear functions and external disturbances. First, the approximation properties of neural networks are used to approximate the MAS partial differential equation (PDE) model with nonlinear terms containing two variables, time t, and spatial variable x. Second, an adaptive controller is constructed to actuate the parabolic MAS to reach consensus under external disturbances. Based on this, the finite-time theorem and special inequalities are applied to prove the stability of the closed-loop system. Thus, MAS that have nonlinear functions and external disturbances are enabled with finite-time consensus. Finally, the effectiveness of the proposed control method is demonstrated by numerical simulations.

Whole Body Control of an Autonomous Mobile Manipulator Using Series Elastic Actuators

The flexible joint robot has superior attractive features because of its high mobility, high load ratiohigh torque fidelity, robustness for external disturbance, task adaptability, and safety. In this paper, an autonomous mobile manipulator driven by the designed series elastic actuators (SEAs) is developed. A complete dynamic model of nonholonomic mobile manipulator including a mobile platform and manipulator with joint flexibility operating simultaneously is proposed. An integrated whole body trajectory control framework is proposed for such robot to perform mobile manipulation tasks. Considering the nonholonomic and holonomic constraints in the mobile manipulation, the whole-body dynamics is formulated and reduced. To address the highly nonlinear of the dynamics and model uncertainty, a novel integral Lyapunov function (ILF)-based adaptive neural network control for task tracking under uncertainties of the flexible joint robot model is proposed. Compared with existing methods, the proposed method provides an alternative for controlling flexible joint robots. The feasibility of the proposed method is verified by the extensive trajectory tracking experiment results in our developed flexible joint manipulator.

Global adaptive neural network control of nonlinear time-varying systems with unknown control coefficients and model uncertainties

This paper investigates the adaptive control problem for the strict-feedback nonlinear time-varying system (NTVS) with unknown control coefficients and model uncertainties. To address the problem that the neural network (NN) can only achieve local approximation of unknown nonlinear functions, a novel NN control scheme is proposed. The scheme consists of a NN controller that operates inside the approximation domain and a robust controller that operates outside the approximation domain, and a novel switching function is designed to enable the two controllers to switch smoothly in the vicinity of the approximation domain to ensure that all signals of the closed-loop system are globally uniformly ultimately bounded. To address the problem that unknown time-varying control coefficients are difficult to handle, an adaptive fault-tolerant control scheme is proposed in combination with the congelation of variables method. The scheme employs classical adaptive control to adapt to the variation of unknown control coefficients and robust control to eliminate the time-varying disturbance terms, and introduces a positive integrable time-varying function to achieve asymptotic tracking of an arbitrary reference signal. The combination of these two schemes constitutes a global adaptive neural network control (GANNC) method for the NTVS with unknown control coefficients and model uncertainties. Finally, an adaptive attitude fault-tolerant controller for launch vehicles is designed by using the GANNC method, which can realize the accurate tracking of the attitude control system to the reference command under normal status, strong disturbances and actuator failures, and comparative experiments prove the effectiveness and superiority of the controller.

Adaptive neural network control of non‐affine multi‐agent systems with actuator fault and input saturation

AbstractAn adaptive neural network fault‐tolerant control method is proposed for the non‐affine multi‐agent systems with actuator failure and input saturation. Through the transformation of the original multi‐agent systems, an equivalent model is proposed. Considering that the signal input of multi‐agent systems is limited, this article will solve the input saturation problem. But in a real system, the actuator would probably fail. Therefore, a fault tolerant control algorithm is proposed to solve the fault problem in the systems. In this article, an adaptive distributed controller is designed to eliminate the uncertainty of the systems. At the same time, the integral barrier Lyapunov function is selected to verify the stability. Finally, a simulation example is given to demonstrate the effectiveness of the proposed method.

Neural network adaptive PID-like coupling control for double pendulum cranes with time-varying input constraints

Industrial cranes commonly experience the double pendulum swing effect as a result of a number of reasons, including the magnitude of the cargo and the non-negligible hook mass. Nevertheless, the majority of control strategies for double-pendulum cranes, whether open-loop or closed-loop, are created using linearized dynamics and necessitate accurate knowledge of the model parameters, rendering them susceptible to parametric uncertainties and unmodeled disturbances. To solve these problems, a new neural network adaptive PID-like coupling control method for the nonlinear double pendulum cranes is proposed in this paper. By making full use of the designed composite signal and its integral action, oscillation suppression and precise positioning are realized. In addition, the problem of conventional PID parameter adjustment is solved successfully by determining the PID gains through a neural network adaptive algorithm, automatically. Moreover, it has been shown that the systematic error converges into a compact set around zero for a variety of time-varying actuator failures and saturation situations that can easily occur in practical engineering. Finally, the effectiveness of the proposed control method is verified through hardware experiments, and the comparison results prove the superior performance of the designed controller.

Vision-admittance-based adaptive RBFNN control with a SMC robust compensator for collaborative parallel robots

This paper proposes a novel vision-admittance-based adaptive neural network control method for collaborative parallel robots tracking a predefined trajectory and force command. First of all, the vision-admittance model is creatively developed for coupling visual information and force-sensing information in the image feature space and is applied to generate the image feature reference trajectory online according to the desired trajectory and the command of the interaction force. Secondly, considering the system modelling uncertainties and external disturbances, an adaptive Radial basis functions neural network (RBFNN) controller is created to realize high-precision trajectory tracking. The adaptive tuning laws are designed based on the Lyapunov stability theorem so that the entire system’s stability and the convergence of the weight adaptation can be guaranteed. Moreover, a sliding mode control (SMC)-based robust compensator is added to the RBFNN controller as an auxiliary term to improve the robustness and stability of the control system. Finally, co-simulations between ADAMS/Simulink are performed on a collaborative Five-bar parallel robot to demonstrate the efficacy of the proposed controller.

Robust adaptive backstepping neural networks fault tolerant control for mobile manipulator UAV with multiple uncertainties

The present study outlines the development of an Adaptive Backstepping Radial Basis Function Neural Networks Fault Tolerant Control (ABRBFNNFTC) methodology. The aforementioned methodology is employed to address the challenge of achieving trajectory following in the context of a Mobile Manipulator Unmanned Aerial Vehicle (MMUAV) when subjected to the effects of actuator faults and parametric uncertainties. The utilization of an adaptive radial basis function neural networks (RBFNNs) controller is employed for the purpose of approximating an unidentified nonlinear backstepping controller that relies on the precise model of the MMUAV. The Lyapunov direct method is utilized to establish the stability analysis of the entire system. The closed-loop system guarantees the Uniformly Ultimately Bounded (UUB) stability of all signals. The control methodology put forth ensures the achievement of a prescribed trajectory, and mitigates the impact of uncertainties and actuator faults. The efficiency of the proposed ABRBFNNFTC scheme is demonstrated through the presentation of extensive simulation studies.

Adaptive Neural Network Control of an Uncertain 2-DOF Helicopter With Unknown Backlash-Like Hysteresis and Output Constraints.

An adaptive neural network (NN) control is proposed for an unknown two-degree of freedom (2-DOF) helicopter system with unknown backlash-like hysteresis and output constraint in this study. A radial basis function NN is adopted to estimate the unknown dynamics model of the helicopter, adaptive variables are employed to eliminate the effect of unknown backlash-like hysteresis present in the system, and a barrier Lyapunov function is designed to deal with the output constraint. Through the Lyapunov stability analysis, the closed-loop system is proven to be semiglobally and uniformly bounded, and the asymptotic attitude adjustment and tracking of the desired set point and trajectory are achieved. Finally, numerical simulation and experiments on a Quanser's experimental platform verify that the control method is appropriate and effective.

Observer-Based Adaptive NN Security Control for Switched Nonlinear Systems Against DoS Attacks: An ADT Approach.

In this article, a novel switched observer-based neural network (NN) adaptive control algorithm is established, which addresses the security control problem of switched nonlinear systems (SNSs) under denial-of-service (DoS) attacks. The considered SNSs are described in lower triangular form with external disturbances and unmodeled dynamics. Note that when an attack is launched in the sensor-controller channel, the controller will not receive any message, which makes the standard backstepping controller not workable. To tackle the challenge, a set of NN adaptive observers are designed under two different situations, which can switch adaptively depending on the DoS attack on/off. Further, an NN adaptive controller is constructed and the dynamic surface control method is borrowed to surmount the complexity explosion phenomenon. To eliminate double damage from DoS attacks and switches, a set of switching laws with average dwell time are designed via the multiple Lyapunov function method, which in combination with the proposed controllers, guarantees that all the signals in the closed-loop system are bounded. Finally, an illustrative example is offered to verify the availability of the proposed control algorithm.

Dynamic surface adaptive RBF neural network control for a class of non‐linear uncertain systems

AbstractAn adaptive neural network dynamic surface control method for strict feedback non‐linear systems with model uncertainty is proposed in this paper. The uncertain parts are approached by RBF neural network; meanwhile, the virtual controller is designed to make system stable according to dynamic surface control. The method of Lyapunov is used to prove the stability and convergence of the system. The simulation results prove the feasibility of this controller and prove that the controller has an advantage to approach the uncertain non‐linear system and make system have well convergence and traceability.

Observer-based distributed convex optimization of bipartite containment control for higher order nonlinear uncertain multi-agent systems

This paper studies the distributed convex optimization of bipartite containment control problem for a class of higher order nonlinear multi-agent systems with uncertain states. For the optimization problem, the penalty function is constructed by summing the local objective function of each agent and combining the penalty term formed by the adjacency matrix. For the unknown nonlinear function and unpredictable states in the system, this paper construct radial basis function Neural-networks and state observer for approaching, respectively. In order to avoid “explosion of complexity,” under the framework of Lyapunov function theory, we propose the dynamic surface control (DSC) technology and design the distributed adaptive backstepping neural network controller to ensure all the signals remain semi-global uniformly ultimately bounded in the closed-loop system and all agents can converge to the convex hull containing each boundary trajectory as well as its opposite trajectory different in sign. Simulation results confirm the feasibility of the proposed control method.

Event‐triggered adaptive neural network control design for stochastic nonlinear systems with output constraint

SummaryThis paper is concerned with the adaptive neural network event‐triggered control (ETC) problem for stochastic nonlinear systems with output constraint. The influence of stochastic disturbance inevitably exists in many practical systems, which leads to system instability. Meanwhile, a novel tan type barrier Lyapunov function (Tan‐BLF) structure is proposed to deal with the constraint requirements of stochastic systems. In the sense of probability, the output constraints will not be violated during the operation of the system. In addition, the ETC strategy is adopted to reduce the burden of communication. The asymptotic stability of the closed‐loop system is guaranteed without violating output constraints. Meanwhile, the tracking error converges to a small region of the origin. Two simulations results demonstrate the effectiveness of theoretical analysis.

Backpropagation Neural Network Adaptive Voltage Control of a High-Gain Transformerless DC/DC Boost Converter for solar applications

An artificial neural network (ANN) adaptive control method for high-gain transformerless DC-DC boost converter is designed in order to achieve better performance and robustness. An ANN controller have the ability to predict the duty cycle in each moment because it is already trained with a vast number of random values of input, output and reference voltages. In order to minimize the error of obtaining the most convenient duty cycle, a back-propagation algorithm has been used to train the neural network because it has the ability to deal with the weights constantly for the sake of diminishing the loss function which makes the ANN generates the fittest duty cycle. The control scheme has been done for a photovoltaic system and the simulation is done using Matlab/Simulink. Simulation results are given for irradiance, reference voltage and load variations. The results are designated to prove the effectiveness of the proposed control system.

Research and analysis of manipulator control method based on deep learning

Aiming at the stability and accuracy of grasping objects, this paper studies the adaptive neural network control and learning of manipulator control system with unknown system dynamics. First, a stable adaptive neural network (NN) controller is designed, and the unknown closed-loop system dynamics of the manipulator is approximated by using radial basis function (RBF) neural network. The LSTM long and short memory algorithm is introduced, and the RBF based hybrid neural network model is constructed. The LSTM algorithm is used to design the input gate, forgetting gate and output gate structures to suppress the gradient expansion problem during coordinate data training, and accurate trajectory correction instructions are given. Lyapunov stability theorem is used to analyze stability. Partial persistent excitation (PE) conditions of some internal signals in the closed-loop system are satisfied in the control process of tracking the circular reference trajectory. Under the PE condition, the proposed adaptive NN controller can accurately identify the dynamic uncertainties of the manipulator in the stability control process. Then, a new NN learning control method is proposed, which effectively uses the knowledge learned without re adapting to the unknown manipulator control system dynamics to achieve closed-loop stability and improve control performance. The effectiveness of the proposed method is studied through simulation.Keyboard

Asymptotical Neuro-Adaptive Consensus of Multi-Agent Systems With a High Dimensional Leader and Directed Switching Topology.

We study the asymptotical consensus problem for multi-agent systems (MASs) consisting of a high-dimensional leader and multiple followers with unknown nonlinear dynamics under directed switching topology by using a neural network (NN) adaptive control approach. First, we design an observer for each follower to reconstruct the states of the leader. Second, by using the idea of discontinuous control, we design a discontinuous consensus controller together with an NN adaptive law. Finally, by using the average dwell time (ADT) method and the Barbǎlat's lemma, we show that asymptotical neuroadaptive consensus can be achieved in the considered MAS if the ADT is larger than a positive threshold. Moreover, we study the asymptotical neuroadaptive consensus problem for MASs with intermittent topology. Finally, we perform two simulation examples to validate the obtained theoretical results. In contrast to the existing works, the asymptotical neuroadaptive consensus problem for MASs is firstly solved under directed switching topology.

Adaptive neural network control of DC–DC power converter

This article proposes a novel Zernike radial neural network based adaptive control architecture for the closed-loop control of output DC voltage in DC-DC buck power converter. The proposed combination of novel Zernike radial neural network estimator and the adaptive backstepping controller effectively compensates for wide range of perturbations affecting the converter system, in an online manner. The closed loop stability of the DC-DC buck power converter with the proposed neuro-adaptive backstepping controller is proven by Lyapunov stability criterion. Numerical simulations are conducted to examine the effectiveness of the proposed controller under start-up response and step changes in the load, source voltage and reference output voltage. Furthermore, the simulation findings are validated by conducting extensive real-time investigation on a laboratory prototype, under a wide range of operating points. The results obtained shows a significant improvement in the transient response of both output voltage and inductor current of the converter, relative to the relevant control methods proposed in the recent past.

Adaptive neuromuscular control of a simplified muscle tendon-driven musculoskeletal system

PurposeThis study aims to improve the muscle model control performance of a tendon-driven musculoskeletal system (TDMS) to overcome disadvantages such as multisegmentation and strong coupling. An adaptive network controller (ANC) with a disturbance observer is established to reduce the modeling error of the musculoskeletal model and improve its antidisturbance ability.Design/methodology/approachIn contrast to other control technologies adopted for musculoskeletal humanoids, which use geometric relationships and antagonist inhibition control, this study develops a method comprising of three parts. (1) First, a simplified musculoskeletal model is constructed based on the Taylor expansion, mean value theorem and Lagrange–d’Alembert principle to complete the decoupling of the muscle model. (2) Next, for this simplified musculoskeletal model, an adaptive neuromuscular controller is designed to acquire the muscle-activation signal and realize stable tracking of the endpoint of the muscle-driven robot relative to the desired trajectory in the TDMS. For the ANC, an adaptive neural network controller with a disturbance observer is used to approximate dynamical uncertainties. (3) Using the Lyapunov method, uniform boundedness of the signals in the closed-loop system is proved. In addition, a tracking experiment is performed to validate the effectiveness of the adaptive neuromuscular controller.FindingsThe experimental results reveal that compared with other control technologies, the proposed design techniques can effectively improve control accuracy. Moreover, the proposed controller does not require extensive considerations of the geometric and antagonistic inhibition relationships, and it demonstrates anti-interference ability.Originality/valueMusculoskeletal robots with humanoid structures have attracted considerable attention from numerous researchers owing to their potential to avoid danger for humans and the environment. The controller based on bio-muscle models has shown great performance in coordinating the redundant internal forces of TDMS. Therefore, adaptive controllers with disturbance observers are designed to improve the immunity of the system and thus directly regulate the internal forces between the bio-muscle models.

Blockchain-enabled peer-to-peer energy trading and resilient control of microgrids

The deregulation and decentralization of the energy market have resulted in a proliferation of distributed generation that participates in energy trading as prosumers. In peer-to-peer (P2P) trading of energy within the microgrid (MG), the peers can trade energy without the need for an intermediary. Blockchain technology is devised to assure the security and resilience of the system's P2P trading against adversarial attacks. The large number of renewable prosumers who participate in trading raises the MG system's oscillation frequency. To regulate the system frequency during trading, a distributed-based federated learned fractional-order recurrent neural network (FL-FORNN) adaptive controller is proposed. The control system is a crucial component of MGs in order to ensure stable performance. To aggregate the network weights, the proposed FL-based controller frequently communicates with the cloud server. To avoid the privacy threat during this case, we further propose to integrate FL with local differential privacy (LDP) to secure against the false data injection attack from the eavesdropper. To validate, the MG model is implemented in OPAL-RT with its resilient controller. The P2P trading of energy in the blockchain is executed in Raspberry Pis (RPis) based on the numerous prosumers/consumers participating in the trading. Then, the power information from the tertiary control implemented in RPis is communicated with the MG secondary frequency controller by interfacing using the user datagram protocol. The proposed work is realized for the MG considering four prosumers and three consumers, and the resiliency of the controller is authenticated with case studies. The results divulge that the LDP of the proposed controller can provide a robust and secure solution of MGs with P2P trading, even in the presence of adversarial attacks.

Distributed Adaptive Optimization Algorithm for Fractional High-Order Multiagent Systems Based on Event-Triggered Strategy and Input Quantization

This paper investigates the distributed optimization problem (DOP) for fractional high-order nonstrict-feedback multiagent systems (MASs) where each agent is multiple-input–multiple-output (MIMO) dynamic and contains uncertain dynamics. Based on the penalty-function method, the consensus constraint is eliminated and the global objective function is reconstructed. Different from the existing literatures, where the DOPs are addressed for linear MASs, this paper deals with the DOP through using radial basis function neural networks (RBFNNs) to approximate the unknown nonlinear functions for high-order MASs. To reduce transmitting and computational costs, event-triggered scheme and quantized control technology are combined to propose an adaptive backstepping neural network (NN) control protocol. By applying the Lyapunov stability theory, the optimal consensus error is proved to be bounded and all signals remain semi-global uniformly ultimately bounded. Simulation shows that all agents reach consensus and errors between agents’ outputs and the optimal solution is close to zero with low computational costs.