Neural Control Scheme Research Articles

Adaptive neural tracking control of stochastic nonaffine nonlinear switched systems with unknown backlash-like hysteresis

This paper is concerned with the tracking control problem for a class of switched nonaffine stochastic nonlinear systems in completely nonaffine form and nonlower-triangular structure, with unknown backlash-like hysteresis involved, and a novel adaptive neural tracking control scheme, based on backstepping design, is proposed. To eliminate the problem of complexity explosion, dynamic surface control (DSC) technique is incorporated into the backstepping design procedure, such that the process of controller design becomes much simpler. High-order neural networks (HONNs) are employed to approximate the lumped unknown nonlinear functions, and only one adaptive parameter is required to be updated. Stability analysis shows that the proposed scheme guarantees all the closed-loop error signals are semi-globally uniformly ultimately bounded in the 4th-moment or mean square, and the system output can converge to an arbitrary small neighbourhood of the given trajectory. Finally, simulation results are presented to verify the effectiveness of the proposed approach.

An IGAP-RBFNN-based secondary control strategy for islanded microgrid-cyber physical system considering data uploading interruption problem

At present, the traditional islanded microgrid is changing to a cyber-physical system. And the data in its cyber layer is often used to solve practical problems in the physical layer through sensors and communicators. But the use of communication data also brings a lot of problems. Among them, the upload interruption problem (i.e., communicator or sensor failure) will make a heavy impact on the system. To improve the voltage and frequency stability of microgrid through secondary control considering data uploading interruption, an improved growing and pruning-radial basis function neural network-based secondary control strategy is designed in this study. The main designs can be summarized as below: (1) When the communicator cannot receive data, a prediction compensation part is designed directly in the corresponding communicator, i.e., the predicted data is used to complete the follow-up works; (2) When the sensor cannot send data, a nonlinear cyber-physical vulnerability assessment part and a neighbor transmission part are designed. In this way, a neighbor sensor is selected to send the interrupted data for the follow-up control process; (3) Based on these designs above, a novel secondary control is designed for microgrid. Finally, the simulation results show the effectiveness of the proposed strategy.

Adaptive neural impedance control for electrically driven robotic systems based on a neuro-adaptive observer

This paper proposes an adaptive neural impedance control (ANIC) strategy for electrically driven robotic systems, considering system uncertainties and external disturbances. For the considered robotic system, the joint velocities and armature currents are assumed to be unknown and unmeasured, and an adaptive observer is then designed to estimate its unknown states using a neural network. Based on the observed joint velocities and armature currents, an ANIC scheme is proposed and the performances of the joint positions and force tracking can be improved. We also prove that the control system is stable and all the signals in closed-loop system are bounded. Simulation examples on a two-link electrically driven robotic manipulator are presented to show the effectiveness of the proposed observer-based intelligent impedance control method.

Neural networks-based adaptive dynamic surface control for vehicle active suspension systems with time-varying displacement constraints

This paper addresses controlling of displacements of vehicle active suspension systems with an active seat suspension and a human-body model. A neural networks-based dynamic surface control strategy is constructed for the active suspension systems. Specifically, asymmetric time-varying barrier Lyapunov functions are applied to ensure that the displacements of vehicle active suspensions do not violate their time-varying constraint bounds. Neural networks are used to approximate unknown functions in the active suspension systems and their basis function properties are employed to deal with functions with non-strict form. Dynamic surface control technique is used to reduce the complexity of the controller. Advantages of the control strategy are substantiated by two simulation examples.

Fully Adaptive-Gain-Based Intelligent Failure-Tolerant Control for Spacecraft Attitude Stabilization Under Actuator Saturation

This article investigates the attitude stabilization problem of a rigid spacecraft with actuator saturation and failures. Two neural network-based control schemes are proposed using anti-saturation adaptive strategies. To satisfy the input constraint, we design two controllers in a saturation function structure. Taking into account the modeling uncertainties, external disturbances, and adverse effects from actuator faults and failures, the first anti-saturation adaptive controller is implemented based on radial basis function neural networks (RBFNNs) with a fixed-time terminal sliding mode (FTTSM) containing a tunable parameter. Then, we upgrade the proposed controller to a fully adaptive-gain anti-saturation version, in order to strengthen the robustness and adaptivity with respect to actuator faults and failures, unknown mass properties, and external disturbances. In the two schemes, all of the designed adaptive parameters are scalars, thus they only require light computational load and can avoid the redesign process of the controller during spacecraft operation. Finally, the feasibility of the proposed methods is illustrated via two numerical examples.

Adaptive recursive sliding-mode dynamic surface and its event-triggered control of uncertain non-affine systems

An adaptive neural network control strategy and its event-triggered controller are designed for non-affine pure-feedback uncertain systems based on nonlinear gains and recursive sliding-mode dynamic surfaces. A kind of nonlinear gain functions is introduced into the traditional framework of dynamic surface control (DSC) with recursive sliding-mode to make a compromise between control accuracy and transient performance. Radial basis function (RBF) neural networks (NNs) are adopted to approximate unknown functions at each step, and novel adaptive update laws with leakage terms of σ-modification is constructed. To reduce the action number of the actuator, an extended control strategy in an event-triggered manner is proposed. By the Lyapunov function, it is proven that both of the two control strategies can force the tracking error arbitrarily small and guarantee all the signals in the closed-loop system uniformly ultimately bounded. Finally, simulation results are provided to verify the effectiveness of the proposed control strategy.

Distributed Adaptive Neural Consensus Control for Stochastic Nonlinear Multiagent Systems with Whole State Delays and Multiple Constraints

This paper presents a distributed adaptive neural tracking consensus control strategy for a class of stochastic nonlinear multiagent systems with whole state time delays, input and output constrains. The considered systems are involved in the existence of whole state delays and stochastic disturbances, which makes the controller design more difficult and complex. Firstly, time delays are related to unknown dynamic interactions with the whole states of the agent systems, and novel Lyapunov-Krasovskii functionals are constructed. Secondly, the smooth asymmetric saturation nonlinearity is given based on Gaussian error function, output constraints are achieved via barrier Lyapunov functions, and neural networks are utilized to deal with the completely unknown nonlinearities and stochastic disturbances. Then, based on Lyapunov stability theory, a delay-independent adaptive controller is developed via Lyapunov-Krasovskii functionals and backstepping technique, and it reduces the complexity of learning parameters. It is proved that the proposed approximation-based controller can guarantee that all closed-loop signals are cooperatively semi-globally uniformly ultimately bounded (CSGUUB), and the tracking errors between the followers and the leaders eventually converge to a small neighbourhood around the origin. Finally, simulation studies are carried out, and the simulation results verify the correctness and effectiveness of the proposed strategy.

Adaptive neuronal induction motor control with an 84‐pulse voltage source converter

AbstractThis paper deals with the problem of harmonic distortion in velocity control drives for large horsepower three‐phase induction motors. A new solution alternative to considerably reduce harmonic distortion in controlled large‐capacity induction motors is introduced. An adaptive neural velocity reference trajectory tracking control scheme based on an 84‐pulse voltage source converter for large horsepower three‐phase induction motors is proposed. Desired flux modulus control is simultaneously performed. Adaptive controller parameters are adjustable online using B‐spline artificial neural networks. Variable load torque of the handled mechanical dynamic system is assumed to be uncertain. Real‐time estimations or measurements of dynamic load torque are unnecessary. Bézier curves are used for desired smooth motion planning to take a large induction motor from its starting towards an operating velocity. Moreover, motion planning is exploited for evasion of harmful mechanical oscillations as well as large peak values of voltages and currents. Closed‐loop efficient velocity profile tracking is confirmed on a 500‐hp induction motor. Comparisons with 6‐pulse and 12‐pulse conventional voltage source converters are also included to highlight the superior energy efficiency of the proposed 84‐pulse ac motor drive. A very low total harmonic distortion of the multiple‐pulse reconstructed three‐phase control voltage signals is also proved. Analytical and numerical results prove the effectiveness and efficiency of the introduced 84‐pulse adaptive neuronal dynamic tracking control strategy.

Adaptive neural network control for stabilizing sphere of floated inertial platform using rotation vector

This paper presents a neural network adaptive control scheme for the sphere three-dimensional stabilization of the floated inertial platform in consideration of the shell rotation and unknown attributions. Firstly, based on the analysis of the moments acting on the sphere, the dynamic model of the rotational sphere is established in view of the unknown attributions including mass distribution, hydrodynamic drag, electric brush friction, disturbances etc. Secondly, a radial basis function neural network is designed to identify the real dynamic model. Thirdly, the rotation vector from current attitude to the reference attitude is selected as the input of the controller according to the Euler rotation theorem. An updating law of the control parameters is derived from the radial basis function neural network. The adaptive controller is designed using this rotation vector and the updating law. This method addresses the problem of different dimensions between the two state variables of the dynamic model, which is derived from the sphere attitude expressed by a quaternion. Finally, simulation examples are provided to verify the effectiveness and robustness of the designed controller.

Adaptive finite‐time tracking control for output‐constrained nonlinear systems with non‐strict‐feedback structure

SummaryThis article investigates the issue of adaptive finite‐time tracking control for a category of output‐constrained nonlinear systems in a non‐strict‐feedback form. First, by utilizing the structural characteristics of radial basis function neural networks (RBF NNs), a backstepping design method is extended from strict‐feedback systems to a kind of more general systems, and NNs are employed to approximate unknown nonlinear functions. In addition, the system output is constrained to the specified region by applying the barrier Lyapunov function (BLF) technique. Furthermore, the finite‐time stability of the system is proved by employing the Bhat and Bernstein theorem. As a result, an adaptive finite‐time tracking control scheme for the output‐constrained nonlinear systems with non‐strict‐feedback structure is proposed by applying RBF NNs, BLF, finite‐time stability theory, and adaptive backstepping technique. It is demonstrated the finite‐time stability of the system, the prescribed convergence of the system output and tracking error, the boundedness of adaptive parameters and state variables. Finally, a simulation example is implemented to illustrate the effectiveness of the presented neural control scheme.

Adaptive Tracking Control of Cooperative Robot Manipulators With Markovian Switched Couplings

Many cooperative robotic systems have not only modeling heterogeneity and uncertainty but also switched couplings, causing control difficulties. Here, we develop a neural network adaptive control framework for cooperative robot manipulators with unknown Euler–Lagrange dynamics and Markovian switched couplings. Second-order Markovian switching networks are used for modeling such cooperative robotic systems, which admit a hybrid neural network control with a desired tracking performance. The hybrid neural network control scheme contains a distributed adaptive controller and a hybrid adaptation law, enabling learning in the closed-loop system. The position and velocity tracking errors are shown to be practically uniformly exponentially stable in the mean-square sense, respectively, guaranteeing the second-order practical tracking. The results also suggest that the neural weight evolves with practical convergence to the ideal, showing the effect of network structure on the adaptation capacity.

Self-adapting J-type air-based battery thermal management system via model predictive control

Battery thermal control plays an indispensable role in terms of the safety and performance for electric vehicles. For air-based cooling technologies, one of the most pressing challenges is to balance the temperature uniformity and constrain the maximum temperature simultaneously under varying driving conditions. This paper proposes a self-adaptive intelligent neural network-based model predictive control strategy for a J-type air-based battery thermal management system. The J-type structure is first optimized through surrogate-based optimization to improve the temperature uniformity before control. Based on the optimized J-type configuration, an operation mode switching module is developed to mitigate the temperature unbalance. The thermal control approach is tested using an integrated driving cycle, and its evaluations are threefold: (i) the neural network-based control without mode switching fails to meet the thermal requirements; (ii) the control with mode switching succeeds in constraining the maximum temperature and maintaining the temperature uniformity within 1.33 K; (iii) the added model predictive control approach slightly enhances the thermal performance but improves the energy efficiency significantly by 15.8%. The results show that the J-type structure with its appropriate control strategy is a promising solution for light-duty electric vehicles using an air-cooling technology.

Adaptive Neural Fault-Tolerant Control for the Yaw Control of UAV Helicopters with Input Saturation and Full-State Constraints

In this paper, an adaptive neural fault-tolerant tracking control scheme is presented for the yaw control of an unmanned-aerial-vehicle helicopter. The scheme incorporates a non-affine nonlinear system that manages actuator faults, input saturation, full-state constraints, and external disturbances. Firstly, by using a Taylor series expansion technique, the non-affine nonlinear system is transformed into an affine-form expression to facilitate the desired control design. In comparison with previous techniques, the actuator efficiency is explicit. Then, a neural network is considered to approximate unknown nonlinear functions, and a time-varying barrier Lyapunov function is employed to prevent transgression of the full-state variables using a backstepping technique. Robust adaptive control laws are designed to handle parameter uncertainties and unknown bounded disturbances to cut down the number of learning parameters and simplify the computational burden. Moreover, an auxiliary system is constructed to guarantee the pitch angle of the UAV helicopter yaw control system to satisfy the input constraint. Uniform boundedness of all signals in a closed-loop system is ensured via Lyapunov theory; the tracking error converges to a small neighborhood near zero. Finally, when the numerical simulations are applied to a yaw control of helicopter, the adaptive neural controller demonstrates the effectiveness of the proposed technique.

Robust formation tracking and collision avoidance for uncertain nonlinear multi-agent systems subjected to heterogeneous communication delays

In this paper, formation control problem for multiple uncertain nonlinear second-order agents in the presence of heterogeneous communication delays is addressed. A continuous repulsive vector is incorporated into agents’ velocity to ensure collision avoidance, whose time derivative is approximated via a finite-time robust integral of the sign of the error (RISE)-like observer. Using adaptive neural control scheme, unknown model dynamics, external disturbances, RISE estimation error, and the time-delays among agents are robustly addressed. The sufficient conditions on the stability of the overall system and collision avoidance are derived using Lyapunov–Krasovskii functionals and algebraic graph theory, and it is proven that the formation tracking error converges to a small neighborhood of zero. Numerical simulation results are also given to illustrate the effectiveness of our proposed methods.

Adaptive neural tracking control for non-affine nonlinear systems with finite-time output constraint

In this paper, an adaptive neural tracking control approach is presented for a class of non-affine nonlinear systems with finite-time output constraint. The non-affine form of the original system is converted to the affine form by means of the mean value theorem. To constraint the output tracking error within a predefined boundary in finite time, a modified performance function, i.e., finite-time performance function, is introduced. During the process of controller design, the dynamic surface control is employed to handle the ‘explosion of complexity’ problem occurred in the conventional backstepping method. According to the approximation of radial basis functions neural networks, an adaptive neural control scheme is developed which guarantees that the output tracking error is preserved within a specified prescribed performance and all the signals in closed-loop system are bounded by appropriately choosing the design parameters. The stability of the closed-loop system is proved by Lyapunov stability analysis and the simulation results are proposed to demonstrate the effectiveness of the developed control approach.

Deterministic learning from neural control for uncertain nonlinear pure‐feedback systems by output feedback

SummaryThe essence of intelligence lies in the acquisition/learning and utilization of knowledge. However, how to implement learning in dynamical environments for nonlinear systems is a challenging issue. This article investigates the deterministic learning (DL) control problem for uncertain pure‐feedback systems by output feedback, which achieves the human‐like learning and control in a simple way. To reduce the complexity of control design and analysis, first, by combining an appropriate system transformation, the original pure‐feedback system is transformed into a simple normal nonaffine system. An observer is then introduced to estimate the transformed system states. Based on the backstepping and dynamic surface control techniques, a simple adaptive neural control scheme is first developed to guarantee the finite time convergence of the tracking error using only one neural network (NN) approximator. Second, through DL, the exponential convergence of the NN weights is obtained with the satisfaction of partial persistent excitation condition. Thus, locally accurate approximation/learning of the transformed unknown system dynamics is achieved and stored as constant NNs. Finally, by utilizing the stored knowledge, an experience‐based controller is constructed and a novel learning control scheme is further proposed to improve the control performance without any further adaptation online for the estimate neural weights. Simulation results have been given to illustrate that the proposed scheme not only can learn and memorize knowledge like humans but also can utilize experience to achieve superior control performance.

Adaptive Neural Tracking Control of Full-state Constrained Nonstrict-feedback Time-delay Systems with Input Saturation

In this study, an adaptive neural backstepping control scheme is proposed for a class of nonstrict-feedback time-delay systems with input saturation, full-state constraints and unknown disturbances. A structural property of radial basis function neural network is presented to deal with the design from the nonstrict-feedback formation. This method does not require the parameter separation technique and its assumption. With the help of the Lyapunov-Krasovskii functionals and Young’s inequalities, the effects of time delays are compensated, and the unknown disturbances are eliminated in the design process. The barrier Lyapunov function (BLF) is applied to arrest the violation of the full-state constraints. To overcome the problem of input saturation nonlinearity, the smooth nonaffme function of the control input signal is adopted to approach the input saturation function. Moreover, an adaptive backstepping neural control strategy is proposed. The proposed adaptive neural controller ensures that all the closed-loop signals are semi-globally uniformly ultimately bounded (SGUUB). Furthermore, the tracking error can converge to a small neighborhood of the origin. The simulation result shows the effectiveness of this method.

Adaptive neural fault-tolerant control for uncertain MIMO nonlinear systems with actuator faults and coupled interconnections

Handling both intermittent actuator faults and coupled interconnections in uncertain multiple-input–multiple-output nonlinear system is still a challenge in the control community. In this paper, to address this issue, an adaptive neural fault-tolerant control scheme is developed. Firstly, neural networks with random hidden nodes are used to approximate unknown functions, and an inequality is introduced to construct controllers such that the singularity problem of the controllers can be circumvented. Secondly, a projection algorithm is adopted to update online the estimated parameters in the controllers such that the boundedness of estimated parameters is ensured. In particular, the boundedness of estimate of unknown fault parameters with intermittent jumps can be definitely guaranteed. Due to the effects of intermittency jumps of unknown parameters on the system stability during operation, a modified Lyapunov function is developed to prove the system stability. It is proved that the system stability only depends on the jumping amplitude of Lyapunov function and the minimum fault time interval and is not affected by the total number of faults. Thirdly, a root mean square type of bound for the tracking error is established by using iterative calculation to illustrate that the system transient performance in the sense of the tracking error is adjustable by proper choice of design parameters. Finally, simulation studies are carried out to verify the effectiveness of the theoretical results.

A Novel Finite-Time Control for Nonstrict Feedback Saturated Nonlinear Systems With Tracking Error Constraint

This article investigates the neural network-based finite-time control issue for a class of nonstrict feedback nonlinear systems, which contain unknown smooth functions, input saturation, and error constraint. Radial basis function neural networks and an auxiliary control signal are adopted to identify unknown smooth functions and deal with input saturation, respectively. The issue of error constraint is solved by combining the performance function and error transformation. Based on the backstepping recursive technique, a neural network-based finite-time control scheme is developed. The developed control scheme can ensure that the closed-loop system is semi-globally practically finite-time stable. Finally, the validity of theoretical results is verified via simulation studies.

Neural Network Predictive Control of Swing Phase for a Variable-Damping Knee Prosthesis With Novel Hydraulic Valve

It is necessary to develop an effective knee prosthesis to recover the lost mobility of amputees. Variable-damping knee prostheses utilizing hydraulic dampers offer several advantages, but current researches are limited in approaching healthy knee behavior. To improve the gait symmetry, this study proposed a variable-damping knee prosthesis with a novel hydraulic damper using neural network predictive control (NNPC) scheme during swing phase. The external fan valve structure of the hydraulic damper can not only realize independent and continuous adjustment of flexion and extension damping by a single motor, but also can effectively avoid the damping adjustment failure caused by excessive load. NNPC was proposed as a controller to control the novel hydraulic damper during swing phase. The online simulation is carried out based on MATLAB utilizing the workbench composed of knee prosthesis prototype and self-built gait simulation and evaluation platform, so as to preliminarily verify the online feasibility and effectiveness of the algorithm at different speeds. The offline gait symmetry experiments are designed to more intuitively and effectively compare the performance of NNPC with the fuzzy logical control proposed in previous work. The results show that NNPC improves the gait symmetry from the fuzzy logical control at different walking speeds as the values of symmetry indices are significantly decrease in the range of 5.13% to 33.96%. These results indicate that the proposed variable-damping knee prosthesis can make a good performance on improving the approximation of healthy gait characteristics and meet the fundamental requirements of walking at various walking speeds.