Neural Control Research Articles

Neural Network Design and Training for Longitudinal Flight Control of a Tilt-Rotor Hybrid Vertical Takeoff and Landing Unmanned Aerial Vehicle

This paper considers a hybrid vertical take-off and landing (VTOL) unmanned aerial vehicle (UAV). By tilting its propellers, the aircraft can transition from rotary-wing (RW) multirotor mode to fixed-wing (FW) mode and vice versa. A novel architecture of a neural network-based controller (NNC) is presented. An “imitative learning” approach is employed to train the NNC to mimic the response of an expert but computationally expensive model predictive controller (MPC). The resulting NNC approximates the MPC’s solution while significantly decreasing the computational cost. The NNC is trained on the longitudinal axis. Successful simulations and real flight tests prove that the NNC is suitable for the longitudinal axis control of a complex nonlinear system such as the tilt-rotor VTOL UAV through a sequence of transitions between the RW mode to the FW mode, and vice versa, in a forward flight.

Adaptive Neural Indirect Inverse Control for a Class of Fractional-Order Hysteretic Nonlinear Time-Delay Systems and Its Application

Adaptive Neural Indirect Inverse Control for a Class of Fractional-Order Hysteretic Nonlinear Time-Delay Systems and Its Application

Neural Learning-Based Adaptive Force Tracking Control for Robots With Finite-Time Prescribed Performance Under Varying Environments

Neural Learning-Based Adaptive Force Tracking Control for Robots With Finite-Time Prescribed Performance Under Varying Environments

Neural Control for Human-Robot Interaction with Human Motion Intention Estimation

Neural Control for Human-Robot Interaction with Human Motion Intention Estimation

Data-based neural controls for an unknown continuous-time multi-input system with integral reinforcement

Data-based neural controls for an unknown continuous-time multi-input system with integral reinforcement

Adaptive neural boundary control for multi-agent manipulators system with uncertainties through cooperative disturbance observers network

This paper addresses vibration control problem of multi-agent flexible manipulators systems in the presence of simultaneous uncertainty and unknown external disturbance. Particularly, the goal is to suppress vibration of both flexible link and joint angular. In this paper, the dynamic model of the considered flexible manipulator is described by the fourth order partial differential equation. Without control, the system is unstable and vibrate constantly due to initial states, the external unknown disturbances and system uncertainties. To compensate the uncertainty in each agent, the neural networks are employed and novel adaptation laws are developed to update weighting parameters in the neural networks. While for the compensation of the external disturbance a cooperative network of disturbance observers is proposed to enhance the observation reliability. With the resulting estimations of uncertainties and the unknown disturbance, adaptive distributed boundary controllers are derived to suppress vibration in-domain and keep joint angular position to zero. The closed-loop system is proven to be uniform ultimately bounded through Lyapunov stability theory. Numerical simulations result shows that compared with the proportional–derivative control, the proposed method almost reduces all overshoot and steady-state error.

Improvement of the Transient Levitation Response of a Magnetic Levitation System Using Hybrid Fuzzy and Artificial Neural Network Control

Improvement of the Transient Levitation Response of a Magnetic Levitation System Using Hybrid Fuzzy and Artificial Neural Network Control

The Extraction Of Neural Strategies From The Surface Emg: 2004-2024.

This review follows two previous papers (Farina et al., 2004, 2014) in which we reflected on the use of surface EMG in the study of the neural control of movement. This series of papers began with an analysis of the indirect approaches of EMG processing to infer the neural control strategies and then closely followed the progress in EMG technology. In this third paper, we focus on three main areas: surface EMG modelling; surface EMG processing, with an emphasis on decomposition; and interfacing applications of surface EMG recordings. We highlight the latest advances in EMG models that allow fast generation of simulated signals from realistic volume conductors, with applications ranging from validation of algorithms to identification of non-measurable parameters by inverse modelling. Surface EMG decomposition is currently an established state-of-the-art tool for physiological investigations of motor units. It is now possible to identify large samples of motor units, to track motor units over multiple sessions, to partially compensate for the non-stationarities in dynamic contractions, and to decompose signals in real-time. The latter achievement has facilitated advances in myocontrol, by using the online decoded neural drive as a control signal, such as in the interfacing of prostheses. Looking back over the 20 years since our first review, we conclude that the recording and analysis of surface EMG signals has seen breakthrough advances in this period. Although challenges in its application and interpretation remain, surface EMG is now a solid and unique tool for the study of the neural control of movement.

Leveraging EMG Signals for Precision Control

This study explores the application of electromyography (EMG) signals for controlling mechanical systems using an Arduino Uno microcontroller, an Olimex EMG shield, sEMG electrodes, and a servo motor. EMG signals have been used for various applications such as prosthetics and assistive devices, proving to be a reliable source for control mechanisms. The experiment involves initial EMG testing, calculating RMS voltage, and integrating servo motor control. Results show that EMG signals can effectively control a servo motor, with improvements achieved through filtering, PID control, and multi-motor setups. Advanced techniques, such as Neural Networks, Reinforcement Learning, and Model Predictive Control, were also explored to enhance system performance. This research demonstrates the potential of EMG-based control systems in robotics and prosthetics, highlighting their future applications and innovations in adaptive technologies.

Adaptive NN Force Loading Control of Electro-Hydraulic Load Simulator

To address the issues of derivative explosion in traditional backstepping control and the strong nonlinearity of hydraulic systems, this paper develops an adaptive neural network control method tailored for electro-hydraulic load simulators. Neural networks are employed to handle external disturbances, modeling uncertainties, and the derivatives of virtual control inputs. First, the precise state-space equations of the system are derived. Next, the approximation property of neural networks is used to design an adaptive backstepping controller, and the symmetric barrier Lyapunov function is used to prove the boundedness of the controller and control parameters. Finally, experiments are conducted to verify the effectiveness and reliability of the control algorithm. The results demonstrate that the proposed control algorithm exhibits excellent tracking performance and effectively reduces control errors.

Research on the Application of Stewart Platform for Wave Compensation

In the field of ocean engineering, the efficiency and safety of ship operation are significantly affected by waves. Faced with the instability and safety risks of ship crane operation under complex sea conditions, a design and research method of efficient wave compensation device based on Stewart platform is proposed in this paper, aiming at developing a compensation system that can respond in real time and effectively counteract the effects of waves, so as to improve the safety and efficiency of offshore operations. Aiming at the limitation of traditional wave compensation technology, this paper adopts the BP neural network PID control algorithm combined with Stewart platform. Utilizing the Stewart platform's six-degree-of-freedom motion capability, fast compensation of wave motion is achieved. On this basis, the BP neural network PID control strategy is introduced to optimize the response speed and compensation accuracy of the system, and realize the efficient and accurate compensation of wave motion under complex sea conditions.

A Fuzzy-Immune-Regulated Single-Neuron Proportional–Integral–Derivative Control System for Robust Trajectory Tracking in a Lawn-Mowing Robot

This paper presents the constitution of a computationally intelligent self-adaptive steering controller for a lawn-mowing robot to yield robust trajectory tracking and disturbance rejection behavior. The conventional fixed-gain proportional–integral–derivative (PID) control procedure lacks the flexibility to deal with the environmental indeterminacies, coupling issues, and intrinsic nonlinear dynamics associated with the aforementioned nonholonomic system. Hence, this article contributes to formulating a self-adaptive single-neuron PID control system that is driven by an extended Kalman filter (EKF) to ensure efficient learning and faster convergence speeds. The neural adaptive PID control formulation improves the controller’s design flexibility, which allows it to effectively attenuate the tracking errors and improve the system’s trajectory tracking accuracy. To supplement the controller’s robustness to exogenous disturbances, the adaptive PID control signal is modulated with an auxiliary fuzzy-immune system. The fuzzy-immune system imitates the automatic self-learning and self-tuning characteristics of the biological immune system to suppress bounded disturbances and parametric variations. The propositions above are verified by performing the tailored hardware in the loop experiments on a differentially driven lawn-mowing robot. The results of these experiments confirm the enhanced trajectory tracking precision and disturbance compensation ability of the prescribed control method.

Fixed-time NN-based adaptive fault-tolerant control for heterogeneous vehicular platoon system with improved exponential spacing policy

This paper is concerned with the distributed fixed-time neural network adaptive tracking control issue for heterogeneous vehicular platoon suffered from actuator faults. Due to the string instability of vehicles caused by non-zero initial spacing errors (ISEs), an improved exponential spacing policy is proposed to eliminate the influence of non-zero ISEs by introducing a compensated term, while guaranteeing the traffic flow stability. Subsequently, a fixed-time platoon control scheme is constructed, where the singularity problem caused by the negative exponent terms of controller is addressed by introducing a cosine-based scalar function. Moreover, the influence of actuator faults and uncertainties for vehicular platoon system is decreased with the control scheme, where the instability caused by maximum acceleration is avoided. The internal stability and string stability of the closed-loop system are proven through Lyapunov stability analysis. The simulation example and comparison results testify to the reliability of control scheme.

Adaptive neural control for a class of random nonlinear Markov jump multi-agent systems with full state constraints

This paper proposes a consensus control protocol based on the adaptive backstepping technique for a class of random Markov jump multi-agent systems (MASs) with full state constraints. Each agent is described by the high-order random nonlinear uncertain system driven by random differential equations, where the random noise is the second-order stationary stochastic process. First, a distributed tracking controller is designed for Markov jump MASs, effectively handling the interaction and coupling terms between agents. Second, an appropriate tan-type barrier Lyapunov function is selected to keep the agents’ states from the violation of constraint boundaries, and the unknown nonlinear terms with Markov jump parameters are approximated by neural networks (NNs) theory. Moreover, to address the differential explosion problem, the extended state observer (ESO) is first introduced instead of employing NNs approximation or command filtering techniques. Finally, the exponentially noise-to-state stability in mean square is proved rigorously through the Lyapunov method, which guarantees all signals of the closed-loop system are bounded in probability and the tracking error converges to a small neighborhood around zero. Simulations are given to demonstrate the feasibility of theoretical results.

Cortical areas for planning sequences before and during movement.

Production of rapid movement sequences relies on preparation before (pre-planning) and during (online planning) movement. Here, we compared these processes and asked whether they recruit different cortical areas. Human participants performed three single-finger and three multi-finger sequences in a delayed movement paradigm while undergoing 7T functional MRI. During preparation, primary motor (M1) and somatosensory (S1) areas showed pre-activation of the first movement, even without increases in overall activation. During production, the temporal summation of activity patterns corresponding to constituent fingers explained activity in these areas (M1 and S1). In contrast, the dorsal premotor cortex (PMd) and anterior superior parietal lobule (aSPL) showed substantial activation during the preparation (pre-planning) of multi-finger compared to single-finger sequences. These regions (PMd and aSPL) were also more active during production of multi-finger sequences, suggesting that pre- and online planning may recruit the same regions. However, we observed small but robust differences between the two contrasts, suggesting distinct contributions to pre- and online planning. Multivariate analysis revealed sequence-specific representations in both PMd and aSPL, which remained stable across both preparation and production phases. Our analyses show that these areas maintain a sequence-specific representation before and during sequence production, likely guiding the execution-related areas in the production of rapid movement sequences.Significance Statement Understanding how the brain orchestrates complex behavior remains a core challenge in human neuroscience. Here, we combine high-resolution neuroimaging and a carefully crafted design to study the neural control of rapid sequential finger movements, like typing or playing the piano. Advancing prior research, we show that the brain areas involved in planning these movements maintain those representations throughout the execution of the sequence. This representational stability across planning and execution suggests an intricate connection between these processes. Our results shed light on the nuanced contributions of different cortical areas to different aspects of coordinating skilled movement. This work is well placed to inform future research in animal models and the development of targeted interventions against movement disorders.

Nonlinear model predictive controller of hydrogenation of dimethyl oxalate for ethylene glycol production

Abstract Ethylene glycol (EG) is a valuable commodity organic intermediate that is produced using the catalyzed gas-phase hydrogenation process of dimethyl oxalate (DMO) from syngas. The reactor process is challenging to control because of its nonlinearity and multivariable condition. Thus, this study proposes the application of Neural Wiener model predictive control (NWMPC) for DMO hydrogenation reactor control. The application of empirical-based MPC, such as NWMPC, is still new in DMO hydrogenation reactor control. In order to simulate the process, the DMO hydrogenation reactor is modeled using Aspen Plus and Aspen Dynamic software. The Neural Wiener (NW) model is developed based on state space and neural network modeling using a Linear-Nonlinear (L-N) identification approach. A validation test is also performed to verify the accuracy of the NW model. Based on the test, the model accuracy is acceptable with the coefficient of determination (R2) of 0.965 for EG output mole fraction (first output) and R2 of 0.936 for product temperature (second output). The NWMPC capability is evaluated with a PID controller to handle a setpoint change in EG output mole fraction and reject disturbance in the feed stream flow rate. The control performance results have demonstrated the superior ability of the NWMPC to handle such scenarios better than PID in terms of controller action speed and profile.

Advanced hybrid control for induction motor combining ANN-PID speed controller with neural predictive current regulator based on particle swarm optimization

This paper presents a hybrid control strategy for an induction motor (IM) organized into two nested loops. In the outer loop, an adaptive artificial neural network-based proportional-integral-derivative (ANN-PID) controller is used for speed control. The ANN-PID is divided into two stages: the first stage includes an adaptive ANN speed estimator, and the second stage includes an adaptation algorithm that fine-tunes both the weights and biases of the ANN estimator and the parameters of the PID controller. In the internal loop, a predictive current controller is used to control IM currents using neural networks. Specifically, the neural predictive controller (NPC) approaches the control problem by presenting it as an optimization problem using a neural predictor for IM currents. The considered objective function includes two elements: the current tracking error and the electromagnetic torque ripple. This function is computed over a given time horizon informed by predictions of IM currents. The particle swarm optimization (PSO) algorithm is applied to determine the optimal solution for this optimization task. The effectiveness of the hybrid control scheme is demonstrated by simulation results obtained using Matlab/Simulink, highlighting its superiority over conventional control methods. The proposed hybrid control is characterized by a reduction in torque ripple and overshoot, while also showing robustness to variations in load, rotor resistance, and rotor inductance. Comparisons with an ANN controller and a fuzzy PI controller further demonstrate the superior performance of the hybrid controller in handling nonlinear dynamics and maintaining control accuracy under various operating scenarios.

Event‐Triggered Adaptive Neural Network Control for 4WS4WD Wheeled Mobile Robot

ABSTRACTIn this article, an event‐triggered adaptive neural network controller based on threshold band is designed for a four wheels independently steered and four wheels independently driven (4WS4WD) mobile robot. The 4WS4WD mobile robot is attracting attention for its excellent motion performance such as manipulation versatility and posture flexibility. However, its control difficulty is increased due to its characteristics of being controlled by eight motors for steering and driving respectively. Also, since the robot itself has limited computing and communication resources, real‐time control cannot be guaranteed. To copy with that, the kinematic and dynamics models are first introduced for the 4WS4WD mobile robot. Second, an adaptive neural network controller with low‐frequency learning rate is utilized to control the mobile robot since there are unknown perturbations in the model. It can maintain system stability while handing unidentified model perturbations. The stability of the controller is demonstrated by Lyapunov stability analysis. An event‐triggered based on threshold band is suggested to lessen the amount of computation in the control process. Finally, the simulation outcomes further demonstrate how the suggested approach can greatly lessen the computational and communication cost while maintaining control performance.

Direct adaptive neural control for precision piezoelectric actuator without hysteresis compensation

This paper introduces a novel Direct Adaptive Neural Control (DANC) approach to enhance the precision of piezoelectric actuator motion control significantly. By effectively addressing the complexities of hysteresis, DANC simplifies system design while delivering superior performance. A Radial Basis Function Neural Network (RBFNN) is employed to handle system uncertainties robustly, and a switching control mechanism guarantees stability, rigorously verified through Lyapunov analysis. Experimental results using a Thorlabs PZS001 actuator convincingly demonstrate substantial performance improvements over conventional PID control, the feedforward-based neural network and feedback control (FF-FB control).

Enhancing Offshore Wind Turbines Performance with Hybrid Control Strategies Using Neural Networks and Conventional Controllers

Abstract A wind turbine is a complex energy converter that requires a robust control solution to overcome performance limitations caused by its mechatronics and external disturbances. Within the turbine operation regions, in the area below the nominal power, the challenge of achieving a precise and highly efficient response must be addressed using multi-objective control strategies that can also contribute to the stability of the structure, reducing vibrations. In this work, a control architecture to maximize the power generation and minimize the vibrations is proposed. Based on this architecture four different hybrid control strategies exploiting radial basis function neural networks and conventional regulators are proposed to cover this twofold objective. The controllers calculate the appropriate electromagnetic torque to track the maximum power point and thus achieve the generation of maximum power, while reducing the acceleration of the tower movement. The neural controllers use a non-supervised learning algorithm to adaptively adjust the weights of the neuros to better couple to the wind turbine dynamics. These combined control methods are tested on a 5 MW floating offshore wind turbine under the influence of winds and waves. The results show that these hybrid strategies are valid for achieving a more efficient response and decreasing turbine fatigue, thus extending their lifetime.