Neural Network Sliding Mode Control Research Articles

Intelligent switching gain based sliding mode control for optimization of power consumption in cooperative manipulators

AbstractCooperative manipulators, vital for intricate tasks, are gaining widespread attention across industries. Recognizing their impact on power consumption, costs, and task outcomes, this paper emphasizes the critical study of control methods, actuator power consumption, and manipulator accuracy. Addressing these challenges, we propose a neural network‐sliding mode controller (NN‐SMC) to optimize actuator power consumption and minimize errors in cooperative manipulators. Operating in trajectory and point‐to‐point modes, the NN‐SMC dynamically generates real‐time switching mode controller (SMC) gains (L and K) for precise control. Stability is ensured by maintaining gains within permissible ranges. In point‐to‐point mode, the NN orchestrates an optimal path generation, along with tailored gains. To evaluate performance accurately, a novel control performance index is introduced. Experimental results on 3‐DOF cooperative manipulators demonstrate a remarkable 28% increase in the control performance index for the trajectory mode and a substantial reduction in computational complexity for both modes. This work not only addresses inherent challenges in cooperative manipulators but also signifies a methodological advancement through the integration of neural network‐based control, promising enhanced efficiency and stability.

Event-triggered adaptive deep neural network sliding mode control design for unmanned aerial vehicle systems

This paper proposes an event-triggered adaptive attitude maneuver control strategy based on deep neural network (DNN) for quadcopters to improve the attitude tracking performance and attenuate the effects of nonlinear and uncertainty. First, we establish an affine function model constructed to describe the dynamic model of quadcopters. The DNN trained over a large time scale is used to approximate the nonlinearity and uncertainty. Then, an adaptive sliding mode control (SMC) is combined with the DNN to achieve accurate tracking of attitude trajectory. Furthermore, to reduce unnecessary data transmissions, an event-trigger mechanism is combined with the designed DNN-driven adaptive SMC. The Lyapunov-based stability analysis is used to prove the stability of the closed loop system. The simulation results show the good performance of the proposed control strategy.

Research on coupling dynamics modeling and composite control of the multi-flexible space robot

Space robots can assist astronauts in accomplishing a variety of space missions in high-risk environments. As spacecraft become increasingly lightweight, large-scale, and integrated, space robots are a type of large-scale robots with flexible links and flexible joints. However, the flexibility of the components complicates robot modeling and controller design. Therefore, to address the issue of stable trajectory tracking for multi-flexible space robots, a dynamics model of space robots with multiple flexibilities is established, and a composite controller is designed to enhance control stability. In modeling, comprehensively considering the rigid joints, joint flexibility and link flexibility, the kinematics is consistently described by using the component deformation matrix and recursive method. An accurate dynamics model is established by Lagrange’s equations. In control, using fuzzy neural network sliding mode control, a rigid subsystem controller is designed to suppress the system jitter caused by sudden parameter changes. And it implements the robust tracking of the robot trajectory. Meanwhile, the weighted average fuzzy controller is designed to control the link deformation. The regulation of the flexible joint by the fast feedback controller is implemented using the virtual torque feedback control. Finally, the validity of the dynamics and control models is verified using Matlab and Adams. Compared with previous studies, this study considers multiple flexibility factors more fully. The vibration of multiple flexible components is suppressed during the stable trajectory tracking of the robot.

Anti-swing control of varying rope length tower crane based on adaptive neural network sliding mode

Due to its complex nonlinear, underdriven, and strongly coupled characteristics, the tower crane will cause the load to swing violently when working, which will cause the tower crane to tip over in serious cases, and there exists a huge safety hazard. In this article, a new artificial neural network sliding mode control method is designed for the control problems of tower crane lifting in position and load swing prevention, which has strong robustness to disturbances and unmodeled dynamics, and ensures that the tower crane lifting is accurately tracked in position and suppresses the load swing at the same time. First, a nonlinear dynamics model of a five-degree-of-freedom tower crane considering the actual working conditions is established. Aiming at the problem that it is difficult to effectively control the nonlinear model of the tower crane system, a new neural sliding mode controller and compensation controller are designed based on the sliding mode control theory and using radial basis function neural network. The neural sliding mode controller is used to approximate the sliding mode equivalent controller with uncertainty and strong nonlinearity, and the compensation controller realizes the compensation of the neural sliding mode controller for the difference between the system control inputs and the uncertainty of the system. The convergence and stability of the proposed control system is rigorously demonstrated using the Lyapunov stability theory. Simulation studies have been carried out to verify the correctness of the model established in this article, as well as the excellent control performance of the control system and the ability to deal with system uncertainty, proving its strong robustness.

Parameter adaptive based neural network sliding mode control for electro‐hydraulic system with application to rock drilling jumbo

SummaryRock drilling jumbo is an important large construction machine used for tunneling construction, and its automation has an urgent demand in engineering. However, the electro‐hydraulic system of the rock drilling jumbo has strong parameters uncertainties and some dynamics that are hard to model accurately, which causes certain challenges for designing model‐based high‐performance control algorithms. To solve these challenges, a parameter adaptive based neural network sliding mode control algorithm is proposed to enhance control performance of the electro‐hydraulic system. The parameter adaptive law is developed to estimate unknown parameters of the system, the neural network is applied for compensating unmodeled dynamics, and then the final control law is designed by sliding mode control method, and the stability demonstration of the closed‐loop system is given. In the simulations, the effectiveness of the designed parameter adaptive law is verified. Extensive comparison experiments are performed on a real rock drilling jumbo driven by proportional valves, the experimental results demonstrate that the developed control algorithm obviously improves the control precision of hydraulic cylinder of the rock drilling jumbo compared with the traditional sliding mode and PID control algorithm, thus the designed control algorithm can be expanded and applied for general hydraulic servo control mechanism.

Design of ABS sliding mode control system for in-wheel motor vehicle based on road identification

Aiming at the problem that the current Anti-lock Braking System (ABS) control algorithm can not make full use of the ground friction to complete the braking when emergency braking on complex roads, an ABS sliding mode control method based on road surface identification is proposed. Combined with the in-wheel motor of in-wheel motor electric vehicle, a coordinated control method of motor regenerative braking and mechanical friction braking is designed. Based on the neural network control method, the road friction coefficient is estimated to realize the identification of typical roads and dynamically obtain the optimal slip ratio of different roads. The ABS sliding mode controler is designed with the optimal slip ratio and the actual slip ratio as input, and the saturation function is used to replace the symbol function in the traditional sliding mode control to weaken the chattering problem, and then the ABS controller is designed. The research results show that compared with the traditional ABS sliding mode controller, the designed ABS neural network sliding mode controller has the following advantages: 1. Through the identification of tire-road friction coefficient by neural network, the real-time performance and road adaptability of ABS controller are improved; 2. When the braking simulation is carried out at a given speed, the braking time is reduced by about 1.03 % and the braking distance is shortened by about 1.01%; 3. The identification of tire-road friction coefficient is realized, which weakens the chattering problem of traditional sliding mode control and improves the stability and robustness of ABS controller.

A Study on the Vehicle Antilock System Based on Adaptive Neural Network Sliding Mode Control

Vehicle antilock systems play a very important role in the stability and reliability during vehicle braking. Due to the complexity of the braking process, antilock braking system (ABS) usually face the problems such as nonlinearity, time-varying, and uncertain parameter modeling. Thus, aiming at the parameter model uncertainty problem of ABS, an adaptive neural network sliding mode controller (ADRBF-SMC) is designed in this paper. On this basis, establishing the quarter-vehicle model and the seven-degree-of-freedom vehicle model, and treating the difference between the two models as a kind of disturbance, carrying out vehicle braking performance simulation experiments to analyze the variation of braking performance parameters such as vehicle and wheel speeds, slip ratio, braking distance, braking torque, under the three cases of adaptive neural network sliding mode controller, traditional sliding mode controller, and no control. Simulation results show that the adaptive neural network sliding mode controller (ADRBF-SMC) proposed in this paper can play an effective control role in both vehicle dynamics models. In addition, the control method proposed in this paper has stronger anti-interference capability and higher robustness compared with the sliding mode controller (SMC).

Integrated Active Suspension and Anti-Lock Braking Control for Four-Wheel-Independent-Drive Electric Vehicles

This paper presents an integrated control scheme for enhancing the ride comfort and handling performance of a four-wheel-independent-drive electric vehicle through the coordination of active suspension system (ASS) and anti-lock braking system (ABS). First, a longitudinal-vertical coupled vehicle dynamics model is established by integrating a road input model. Then the coupling mechanisms between longitudinal and vertical vehicle dynamics are analyzed. An ASS-ABS integrated control system is proposed, utilizing an H∞ controller for ASS to optimize load transfer effect and a neural network sliding mode control for ABS implementation. Finally, the effectiveness of the proposed control scheme is evaluated through comprehensive tests conducted on a hardware-in-loop (HIL) test platform. The HIL test results demonstrate that the proposed control scheme can significantly improve the braking performance and ride comfort compared to conventional ABS control methods.

Adaptive radial basis function neural network sliding mode control of robot manipulator based on improved genetic algorithm

ABSTRACT Since the trajectory-tracking control performance of multi-joint robot manipulator may be degraded due to modeling errors and external disturbances, this paper designs a new adaptive robot manipulator trajectory tracking control method through improved genetic algorithm and radial basis function neural network sliding mode control (IGA-RBFNNSMC). Firstly, the genetic algorithm (GA) is improved by establishing superior populations centered on individuals with high fitness values and selecting individuals in the superior populations for crossover and variation. Secondly, the improved genetic algorithm (IGA) is used for the optimization of the center vector and width vector of the Gaussian basis function in radial basis function (RBF) neural network. Then, based on the dynamics model of the robot manipulator, the modeling errors are approximated by RBF neural network and eliminated by sliding mode control (SMC), and the Lyapunov theorem is used to prove the stability and convergence of the control system. Finally, a two-joint robot manipulator is taken as the research objective and the simulation results show that IGA can significantly reduce the solution time on the basis of guaranteed accuracy and IGA-RBFNNSMC can make the trajectory tracking control accurate and more efficient, which proves the effectiveness of the proposed control method.

Adaptive neural network sliding mode controller design for load following of nuclear power plant

The power control and load following of nuclear reactors have always been a research topic of concern in the field of nuclear power. In order to improve the control system performance and load following performance, this paper proposes a sliding mode control method based on adaptive RBF neural network (RBFNN) for pressurized water reactor (PWR) power control and load following. Firstly, based on the operation mechanism and energy transport process of PWR, the PWR power model is established based on the point-reactor equation. Then, based on Lyapunov stability theory, an ideal sliding mode controller is designed for a general nonlinear second-order model. Secondly, considering that some state variables in the actual PWR model cannot be directly measured, which leads to the problem that the actual control law cannot be directly applied, RBFNN is used as the controller, and its output is used to replace the actual control law and approximate the ideal control law. Adaptive algorithm is adopted to improve the approximation effect of RBFNN. Finally, the simulation results show that the proposed control method can achieve the PWR load following task well under the set working conditions, and has good robustness to disturbances. Compared with the traditional sliding mode control (TSMC) method, this method can effectively eliminate chattering phenomenon and greatly improve the control performance of the sliding mode controller.

Dynamic modeling and rotation control for flexible single-link underwater manipulator considering flowing water environment based on modified morison equation

Flexible single-link underwater manipulator (FSLUM) can be installed in autonomous underwater vehicles, it can carry out underwater tasks such as offshore oil and gas exploration, underwater cable maintenance as well as fishery farming and fishing. The design and development of the FSLUM are of great significance for ocean exploration and development. The FSLUM is affected by the underwater interference torque during underwater tasks in the flowing water environment, which will greatly affect the operational accuracy. In addition, many nonlinear factors such as the flexibility of the transmission system and the two-dimensional deformation of flexible links will aggravate the control difficulty of the FSLUM. An improved sliding mode control (SMC) strategy based on neural network identification is proposed to enhance the FSLUM's control accuracy. Neural networks are proposed to identify uncertain components including underwater interference and flexible nonlinear terms, so as to advance the control law design accuracy. Firstly, the FSLUM dynamic equations coupled with underwater interference torque considering multiple nonlinear factors are established according to the assumed mode method (AMM). A modified Morison equation is proposed to abbreviate the underwater interference torque in flowing water environments. Next, the control law of improved neural network sliding mode control (NNSMC) strategies are proposed according to the Lyapunov stability theorem, so as to ensure closed-loop stability of the FSLUM. Finally, the results of simulation and prototype tracking control experiments show that the improved NNSMC strategy has high tracking accuracy.

Dynamic modeling and neural network compensation for dual-flexible servo system with an underactuated hand

This paper investigates a difficult problem of nonlinear dynamics and motion control of a dual-flexible servo system with an underactuated hand (DFSS-UH). Variation in grasping mass and nonlinear factors of the DFSS-UH including complex flexible deformation and friction torque aggravate the output speed fluctuation, leading to modeling errors in the dynamics, which in turn affects the underactuated hand motion accuracy. A novel neural network sliding mode control (NNSMC) method is designed to control the DFSS-UH. The strategy utilizes neural networks to compensate for dynamics modeling errors, which takes into account neglected nonlinear factors and inaccurate friction torque. The reaching law with the hyperbolic tangent function is proposed to improve sliding mode control, thereby weakening the chattering phenomenon. First of all, the DFSS-UH mechanical model considering many nonlinear factors is established and a dynamic simplification model which ignores higher-order modes is proposed. Secondly, the adaptive law of weighted coefficients is proposed according to the stability of the DFSS-UH. Finally, the physical control platform of the DFSS-UH is built, and simulation and control experiments are conducted. Experimental results show that the improved NNSMC strategy decreases the tracking error of flexible load, thereby enhancing the control accuracy of the DFSS-UH.

A robust voltage control of dual active full bridge converter based on RBF neural network sliding mode control with reduced order modeling approach

This paper presents a control method of dual active full bridge (DAB) DC/DC converter based on neural network Sliding mode control under the reduced order modeling method, which is used to enhance the output voltage regulation and improve the system robustness. Sliding mode control (SMC) is famous for its robustness and improving the dynamic performance of nonlinear systems. However, it includes drawbacks such as complex modeling, chattering, and decreased tracking performance. Therefore, through the method of reduced order modeling, the Radial Basis Function neural network algorithm is selected to modify and design the traditional sliding mode variable structure controller, completing the design of neural network approximation terms and adaptive laws. The proposed reduced order modeling and RBF(Radial Basis Function Neural Network)-SMC scheme simplifies the dual active full bridge modeling process, completes the parameter approximation of the sliding mode controller, and eliminates the chattering problem. Firstly, through simulation and comparative analysis of commonly used modeling methods for dual active full bridge, a reduced order modeling method was adopted to simplify the design process. Then, by ingeniously designing the sliding mode surface and Radial Basis Function control law, and using neural network to modify the Sliding mode control technology, the chattering problem, load disturbance and voltage fluctuation influence of the sliding mode controller are improved. The stability of the control method is proved by Lyapunov stability theory. Finally, the proposed RBF-SMC method is compared with PI linear control and classical Sliding mode control methods through simulation and experiments to verify the effectiveness of the proposed control method.

Adaptive neural network sliding mode control for serially connected hydraulic cylinders of a heavy-duty hydraulic manipulator

Adaptive neural network sliding mode control for serially connected hydraulic cylinders of a heavy-duty hydraulic manipulator

Self-Feedback Neural Network Sliding Mode Control With Extended State Observer for Active Power Filter

In this paper, a fuzzy neural network adaptive sliding mode control with a self-feedback recursion (FNNASMCSFR) based linear extended state observer (LESO) is proposed for a single-phase active power filter (APF), where the adaptive sliding mode controller is designed to improve the response and accuracy of current compensation and reference current tracking. The LESO is designed to estimate the actual APF system dynamics which includes the parameter perturbation and external disturbance. Moreover, the fuzzy neural network with self-feedback recursion is adopted to mimic the switching control gain of ASMC, which combines the output values of neurons at the current time and the previous time, to achieve better dynamic approximation effect and prevent sudden changes. Simulation and hardware experiments verify the introduced method is a viable control solution in harmonics suppression and current control.

Dynamic modeling and neural network compensation for rotating Euler-Bernoulli beam using a novel deformation description method

Dynamic models that consider multiple nonlinear factors including the two-dimensional deformation and higher-order modes are too computationally complex to be suitable for real-time control of the rotating Euler-Bernoulli beam. This article proposes a new dynamic modeling method according to the assumed mode method. The new modeling method not only has high modeling accuracy but also has a brief mathematical expression. In addition, based on the new modeling method, a neural network sliding mode control strategy using the saturation function is designed to improve the rotational accuracy of the Euler-Bernoulli beam. In the control law design, the saturation function is proposed instead of the unit step function to reduce the chattering phenomenon. Simulation and control experiments describe that the new dynamic modeling method has higher modeling accuracy than the modeling method based on the one-dimensional deformation; the proposed strategy can effectively decrease the rotation angle tracking error of the rotating Euler-Bernoulli beam and improve the control accuracy.

Adaptive neural network independent joint-based control for an ODE-PDE rigid-flexible manipulator with multiple constraints

This paper deals with the adaptive neural network control of a nonlinear rigid-flexible manipulator subject to external disturbances and multiple constraints. The system dynamic is modeled by ordinary and partial differential equation (ODE and PDE) based on Hamilton’s Principle. An adaptive neural network sliding mode control strategy is designed to estimate the unknown disturbances and uncertain parameter. The sliding surface consisting of angular error, angular velocity error, and end-point vibration information is proposed based on the energy analysis of the flexible beam. The tangent barrier Lyapunov function is used to ensure the angle errors and end-point deflection within the restrictions. It is shown that the joint motion and vibration of the flexible beam can be adjusted only with an independent joint control input, and no boundary force input is required for vibration suppression. The stability is obtained by semi-group theory and LaSalle’s Invariance Principle. Simulation results demonstrate the effectiveness of the control strategy.

Adaptive neural network sliding mode control for reliable stabilization of uncertain Takagi–Sugeno fuzzy systems regulated by switching rules

The problem of adaptive sliding mode control for a class of continuous-time Takagi–Sugeno fuzzy systems regulated by the event of switching rules relying on neural network estimation method is put forward in this paper, where the plant suffers from state delay, internal structure uncertainty, and unknown nonlinearity. By proposing a switching surface in integral type, it obtains a sliding motion with desired property. In addition, to compensate the plant unknown nonlinearity and to meet the reaching condition, a radial basis function neural-network-based adaptive law is designed to ensure the existence of sliding motion in finite time. Furthermore, for the purpose of exponential stabilization of the sliding motion, a linear matrix inequality condition accompanied with switching signal characterized by an average dwell time is put forward. Finally, two numerical examples, one with all subsystems unstable and the other with stable subsystems and unstable systems, are shown to confirm the validity.

Attitude Fault-Tolerant Control of Aerial Robots with Sensor Faults and Disturbances

In this paper, sensor fault diagnosis and fault tolerant control strategy are investigated for quadcopters under sensor faults and disturbances. We propose the fault diagnosis estimation system and the fault-tolerant control (FTC) method. The fault diagnosis system provides time-varying sensor fault estimation under an unknown bound of disturbances. Moreover, the fault-tolerant control eliminates disturbance that is estimated through the associated disturbance observer. Overall, the proposed FTC guarantees the finite-time tracking convergence using nonsingular fast terminal sliding mode algorithm. Stability of the closed-loop system is validated through the Lyapunov theory. Finally, conventional nonsingular fast terminal sliding mode and adaptive neural network sliding mode control are compared with the proposed method through simulations under sensor faults and disturbances with different scenarios.

Gated Recurrent Fuzzy Neural Network Sliding Mode Control of a Micro Gyroscope

This paper proposes a non-singular fast terminal sliding mode control (NFTSMC) method for micro gyroscopes with unknown uncertainty based on gated recurrent fuzzy neural networks (GRFNNs). First, taking advantage of non-singular fast terminal sliding control, a sliding hyperplane is designed with a nonlinear function to ensure that the tracking error of the system converges to zero within a specified finite time. Then, the unknown model parameters of the micro gyroscope are estimated using a GRFNN. Since the GRFNN can adaptively adjust the base width, center vector, gated recurrent unit parameters, and outer gains, it can achieve accurate approximation to unknown models, enhancing the robustness and accuracy. In addition, due to the introduction of gated recurrent units, The GRFNN can effectively utilize the previous data and avoid the problem of gradient disappearance. The comparison of the simulation results with traditional neural sliding mode control shows that the proposed method can achieve better tracking performance and more accurate estimation of unknown models.