Fuzzy neural networks for direct adaptive control

It is well known that sliding-mode control is simple and insensitive to uncertainties and disturbances. However, control input chattering is the main problem of the classical sliding-mode controller (SMC). In this paper, a fuzzy neural network SMC (FNNSMC) is presented for a class of nonlinear systems. The FNNSMC can eliminate the chattering, unlike the SMC, but there is larger rising time in the FNNSMC than in the SMC. In some cases, small rise time is important. To decrease the rising time of the FNNSMC, an adaptive controller is proposed where the SMC and the FNNSMC are incorporated by a smooth transformation. This adaptive control scheme can improve the dynamical performance and eliminate the high-frequency chattering in the control signal. The system stability is proved by using the Lyapunov function. The simulation results demonstrate the advantages of the proposed adaptive controller.     

Neural-Network-Based Terminal Sliding-Mode Control of Robotic Manipulators Including Actuator Dynamics

A neural-network-based terminal sliding-mode control (SMC) scheme is proposed for robotic manipulators including actuator dynamics. The proposed terminal SMC (TSMC) alleviates some main drawbacks (such as contradiction between control efforts in the transient and tracking errors in the steady state) in the linear SMC while maintains its robustness to the uncertainties. Moreover, an indirect method is developed to avoid the singularity problem in the initial TSMC. In the proposed control scheme, a radial basis function neural network (NN) is adopted to approximate the nonlinear dynamics of the robotic manipulator. Meanwhile, a robust control term is added to suppress the modeling error and estimate the error of the NN. Finite time convergence and stability of the closed loop system can be guaranteed by Lyapunov theory. Finally, the proposed control scheme is applied to a robotic manipulator. Experimental results confirm the validity of the proposed control scheme by comparing it with other control strategies.      

Design of a perturbation estimator using the theory ofvariable-structure systems and its application to magnetic levitationsystems

A perturbation estimator using the theory of variable structure systems is proposed to enhance the robustness of a pole-placement controller design. In its ideal form, the pole-placement design using feedback-linearization technique achieves a desired performance in nonlinear time-varying systems. However, its performance deteriorates rapidly with the presence of disturbance and parametric uncertainties, referred to as perturbation. The estimate generated by the proposed perturbation estimator is incorporated as an additional input to rectify the uncertainties in the nominal control model of the pole-placement design. The proposed scheme requires neither the measurement of the time derivative of the state vector nor the precise knowledge of system parameters, hut rather the bounds on system perturbation. Chatter and the adverse effects of conservative bounds on system perturbation, often encountered in conventional sliding-mode control (SMC), are alleviated for the controlled plant by the proposed scheme. The benefits of this scheme are demonstrated in this study practically on a magnetic levitation system and its performance is compared with that of the conventional SMC scheme     

Analysis and real-time implementation of a radial-basis-function neural-network compensator for high-performance robot manipulators

System performance of robot manipulators with nonadaptive controllers might degrade significantly in the presence of structured or unstructured uncertainties. In order to improve the system performance, a novel radial-basis-function (RBF) neural-network (NN) compensator is proposed. With the RBF NN compensator introduced, the system errors and the NN weights with large dispersion in the initial NN weights are guaranteed to be bounded in the Lyapunov sense. The NN weights of the RBF NN compensator are adaptively tuned. Several software-based controllers, including the computed-torque control (CTC) and a few RBF NN schemes, are implemented in an industrial manipulator in real time. Experimental results are obtained to demonstrate the relative effectiveness of the proposed controllers in improving the tracking performance of the robot manipulators associated with structured or unstructured uncertainties.     

Robust fuzzy neural network sliding mode control scheme for IPMSM drives

This article proposes a robust fuzzy neural network sliding mode control (FNNSMC) law for interior permanent magnet synchronous motor (IPMSM) drives. The proposed control strategy not only guarantees accurate and fast command speed tracking but also it ensures the robustness to system uncertainties and sudden speed and load changes. The proposed speed controller encompasses three control terms: a decoupling control term which compensates for nonlinear coupling factors using nominal parameters, a fuzzy neural network (FNN) control term which approximates the ideal control components and a sliding mode control (SMC) term which is proposed to compensate for the errors of that approximation. Next, an online FNN training methodology, which is developed using the Lyapunov stability theorem and the gradient descent method, is proposed to enhance the learning capability of the FNN. Moreover, the maximum torque per ampere (MTPA) control is incorporated to maximise the torque generation in the constant torque region and increase the efficiency of the IPMSM drives. To verify the effectiveness of the proposed robust FNNSMC, simulations and experiments are performed by using MATLAB/Simulink platform and a TI TMS320F28335 DSP on a prototype IPMSM drive setup, respectively. Finally, the simulated and experimental results indicate that the proposed design scheme can achieve much better control performances (e.g. more rapid transient response and smaller steady-state error) when compared to the conventional SMC method, especially in the case that there exist system uncertainties.     

Adaptive Sliding Mode Control of Robot Manipulator Based on Second Order Approximation Accuracy and Decomposed Fuzzy Compensator

A kind of adaptive sliding mode control scheme for tracking control of robot manipulator which has structured uncertainties and unstructured uncertainties is proposed in this paper. Multi-input Multi-output fuzzy logical system (FLS) is used as a compensator in the control law to compensate all the uncertainties. To reduce the number of the fuzzy rules and the burden of the computation, we design FLS based on second order approximation theorem which can approximate the uncertain function with less fuzzy rules at arbitrary precision than traditional FLS. Besides, to further reduce the number of the fuzzy rules and the amount of calculation, a new decomposed fuzzy logical system based on the decomposition of membership function is proposed. From the simulation results we can see that the control scheme and the fuzzy compensator proposed in this paper can perform fairly.     

Congestion Control for DiffServ Network Using Second-Order Sliding Mode Control

In this paper, a nonlinear fluid flow model is used to analyze and control DiffServ Network. The controller design is based on the integrated dynamic congestion control strategy and a leader–follower control scheme. With respect to standard sliding mode control (SMC), the second-order sliding mode technique shows the same properties of robustness to uncertainties of model and considerable simplification of model used in the design. Apart from the robustness feature, the proposed second-order SMC laws have the advantage of being continuous, thus eliminating the chattering effect and being more acceptable in application. The performance of the control scheme is verified by the simulation results.      

A Type-2 Fuzzy Switching Control System for Biped Robots

In this paper, a type-2 fuzzy switching control system is proposed for a biped robot, which includes switched nonlinear system modeling, type-2 fuzzy control system design, and a type-2 fuzzy modeling algorithm. A new switched system model is proposed to represent the continuous-time dynamic and discrete-event dynamic of a walking biped as a whole, which is helpful to analyze the closed-loop stability of the biped locomotion. A type-2 fuzzy switching control system is proposed for the switched system model to guarantee the gait stability and to achieve a robust control performance with a simplified control scheme. Finally, we propose a new fuzzy c-mean variance algorithm for the type-2 fuzzy system modeling to capture the variance of each clustering means, which can translate random uncertainties of original data into rule uncertainties. Simulation results are reported to show the performance of the proposed control system model and algorithms.     

Leader‐Following Formation Control of Multiple Robots with Uncertainties through Sliding Mode and Nonlinear Disturbance Observer

This paper presents a control scheme for the leader‐following formation of multiple robots. The control scheme combines the sliding mode control (SMC) method with the nonlinear disturbance observer (NDOB) technique. The formation dynamics suffer from uncertainties because the individual robots are uncertain. Concerning such formation uncertainties, the leader‐following formation dynamics are modeled. Assuming that the formation uncertainties have an unknown boundary, an NDOB‐based observer was designed to estimate the formation uncertainties. A sliding surface containing the observer outputs has been defined. Regarding the sliding surface, an SMC‐based controller was investigated to form uncertain robots. A sufficient condition in the sense of the Lyapunov theory was proven such that the formation system is asymptotically stable. Herein, some comparison results between the sole SMC method and the second‐order SMC method are presented to demonstrate the effectiveness and feasibility of the control scheme for multiple robots in the presence of uncertainties.     

Intelligent voltage control strategy for three-phase UPS inverters with output LC filter

This paper presents a supervisory fuzzy neural network control (SFNNC) method for a three-phase inverter of uninterruptible power supplies (UPSs). The proposed voltage controller is comprised of a fuzzy neural network control (FNNC) term and a supervisory control term. The FNNC term is deliberately employed to estimate the uncertain terms, and the supervisory control term is designed based on the sliding mode technique to stabilise the system dynamic errors. To improve the learning capability, the FNNC term incorporates an online parameter training methodology, using the gradient descent method and Lyapunov stability theory. Besides, a linear load current observer that estimates the load currents is used to exclude the load current sensors. The proposed SFNN controller and the observer are robust to the filter inductance variations, and their stability analyses are described in detail. The experimental results obtained on a prototype UPS test bed with a TMS320F28335 DSP are presented to validate the feasibility of the proposed scheme. Verification results demonstrate that the proposed control strategy can achieve smaller steady-state error and lower total harmonic distortion when subjected to nonlinear or unbalanced loads compared to the conventional sliding mode control method.     

Adaptive sliding controller with self-tuning fuzzy compensation for vehicle suspension control

Since the hydraulic actuating suspension system has nonlinear and time-varying behavior, it is difficult to establish an accurate dynamic model for a model-based sliding mode control design. Here, a novel model-free adaptive sliding controller is proposed to suppress the position oscillation of the sprung mass in response to road surface variation. This control strategy employs the functional approximation technique to establish the unknown function for releasing the model-based requirement. In addition, a fuzzy scheme with online learning ability is introduced to compensate the functional approximation error for improving the control performance and reducing the implementation difficulty. The important advantages of this approach are to achieve the sliding mode controller design without the system dynamic model requirement and release the trial-and-error work of selecting approximation function. The update laws for the coefficients of the Fourier series functions and the fuzzy tuning parameters are derived from a Lyapunov function to guarantee the control system stability. The experimental results show that the proposed control scheme effectively suppresses the oscillation amplitude of the vehicle sprung mass corresponding to the road surface variation and external uncertainties, and the control performance is better than that of a traditional model-based sliding mode controller.     

Trajectory control of robotic manipulators using neural networks

The authors present a nonlinear compensator using neural networks for trajectory control of robotic manipulators. The neural networks are not used to learn inverse-dynamics but to compensate nonlinearities of robotic manipulators. The performance of the proposed neural network controller is compared with that of the adaptive controller proposed by J.J. Craig (1988), and the effectiveness of the proposed neural network controller in compensating the unstructured uncertainties is clarified. A learning scheme using a model of known dynamics of manipulators is also proposed. The model learning can be done offline and needs no data recording of actual manipulator operation     

T–S Model-Based SMC Reliable Design for a Class of Nonlinear Control Systems

This paper studies the robust reliable control issues based on the Takagi–Sugeno (T–S) fuzzy system modeling method and the sliding-mode control (SMC) technique. The combined scheme is shown to have the merits of both approaches. It not only alleviates the online computational burden by using the T–S fuzzy model to implement the original nonlinear system (since most of the system parameters of the T–S model can be offline computed) but also preserves the advantages of the SMC schemes, including rapid response and robustness. Moreover, the combined scheme does not require online computation of any nonlinear term of the original dynamics, and the increase in the partition number of the region of premise variables does not create extra online computational burdens for the scheme. Under the design, the control mission can continue safely without prompt external support, even when some of the actuators fail to operate. Meanwhile, both the active and the passive reliable designs are presented. The proposed analytical results are also applied to the attitude control of a spacecraft. Simulation results demonstrate the benefits of the proposed scheme.      

PID-type sliding mode-based adaptive motion control of a 2-DOF piezoelectric ultrasonic motor driven stage

This paper presents a new scheme of adaptive sliding mode control (ASMC) for a piezoelectric ultrasonic motor driven X–Y stage to meet the demand of precision motion tracking while addressing the problems of unknown nonlinear friction and model uncertainties. The system model with Coulomb friction and unilateral coupling effect is first investigated. Then the controller is designed with adaptive laws synthesized to obtain the unknown model parameters for handling parametric uncertainties and offsetting friction force. The robust control term acts as a high gain feedback control to make the output track the desired trajectory fast for guaranteed robust performance. Based on a PID-type sliding mode, the control scheme has a simple structure to be implemented and the control parameters can be easily tuned. Theoretical stability analysis of the proposed novel ASMC is accomplished using a Lyapunov framework. Furthermore, the proposed control scheme is applied to an X–Y stage and the results prove that the proposed control method is effective in achieving excellent tracking performance.     

Analogue implementation of a neural network controller for UPSinverter applications

An analogue neural-network controller for UPS inverter applications is presented. The proposed neural-network controller is trained off-line using patterns obtained from a simulated controller, which had an idealized load-current-reference. Simulation results show that the proposed neural-network controller can achieve low total harmonic distortion under nonlinear loading condition and good dynamic responses under transient loading condition. To verify the performance of the proposed NN controller, a hardware inverter with an analogue neural network (NN) controller (using mainly operational amplifiers and resistors) is built. Additionally, for comparison purposes, a PI controller with optimized parameters is built. Experimental results confirm the simulation results and show the superior performance of the NN controller especially under rectifier-type loading condition. Implementing the analogue neural-network controller using programmable integrated circuits is also discussed     

A Neuro-Sliding-Mode Control With Adaptive Modeling of Uncertainty for Control of Movement in Paralyzed Limbs Using Functional Electrical Stimulation

During the past several years, several strategies have been proposed for control of joint movement in paraplegic subjects using functional electrical stimulation (FES), but developing a control strategy that provides satisfactory tracking performance, to be robust against time-varying properties of muscle-joint dynamics, day-to-day variations, subject-to-subject variations, muscle fatigue, and external disturbances, and to be easy to apply without any re-identification of plant dynamics during different experiment sessions is still an open problem. In this paper, we propose a novel control methodology that is based on synergistic combination of neural networks with sliding-mode control (SMC) for controlling FES. The main advantage of SMC derives from the property of robustness to system uncertainties and external disturbances. However, the main drawback of the standard sliding modes is mostly related to the so-called chattering caused by the high-frequency control switching. To eliminate the chattering, we couple two neural networks with online learning without any offline training into the SMC. A recurrent neural network is used to model the uncertainties and provide an auxiliary equivalent control to keep the uncertainties to low values, and consequently, to use an SMC with lower switching gain. The second neural network consists of a single neuron and is used as an auxiliary controller. The control law will be switched from the SMC to neural control, when the state trajectory of system enters in some boundary layer around the sliding surface. Extensive simulations and experiments on healthy and paraplegic subjects are provided to demonstrate the robustness, stability, and tracking accuracy of the proposed neuroadaptive SMC. The results show that the neuro-SMC provides accurate tracking control with fast convergence for different reference trajectories and could generate control signals to compensate the muscle fatigue and reject the external disturbance.     

Modeling,control and experimental investigation of horizontal hydraulic flight motion simulator

The horizontal hydraulic flight motion simulator (HHFMS) is widely applied in the hardware-in-the-loop (HWIL) simulation of the aircraft attitude attributing to high dynamic response and large power density when a heavy load is tested. In order to achieve a high-precision control performance of the HHFMS, some serious mismatched uncertainties consisting of nonlinear friction torque, unbalanced gravity torque, inertia variation and unmodeled dynamics have to be taken into account. In particular, gravity torque, as an asymmetrical load, will degrade the control performance at the starting moment. In this paper, via transforming mismatched uncertainties into matched uncertainties as a unified disturbance, a cascaded model was firstly established, which can not only avoid designing the complex virtual control laws but also help indirectly reduce the asymmetrical effect of gravity torque. Then, a linear extended state observer (LESO) based continuous sliding mode control (SMC) was proposed. LESO is expected to realize a good suppression of disturbance, as well as convenient acquisition of states signals such as velocity and acceleration. In addition to ensuring the control accuracy and the robustness, continuous SMC without sign function also frees from worrying about possible chattering. Moreover, a setting criterion of two parameters that satisfy Hurwitz's Condition in the third-order SMC was also provided. Finally, experimental investigation shows the effectiveness of dynamic modeling and the practicability of the proposed control method.     

A two-layered fuzzy logic controller for systems with deadzones

Existing fuzzy control methods do not perform well when applied to systems containing nonlinearities arising from unknown deadzones. In particular, we show that a usual "fuzzy PD" controller applied to a system with a deadzone suffers from poor transient performance and a large steady-state error. In this paper, we propose a novel two-layered fuzzy logic controller for controlling systems with deadzones. The two-layered control structure consists of a fuzzy logic-based precompensator followed by a usual fuzzy PD controller. Our proposed controller exhibits superior transient and steady-state performance compared to usual fuzzy PD controllers. In addition, the controller is robust to variations in deadzone nonlinearities. We illustrate the effectiveness of our scheme using computer simulation examples.<>     

Multimode vibration control of a flexible structure usingH∞-based robust control

This paper presents the experimental results of a robust control scheme to suppress the vibration of a flexible structure. The feedback controller is designed using the H_∞-based robust control theory. For this purpose, a flexible bridge tower connected with a crane structure is considered to control its first five vibration modes using a static state feedback controller. A five-degrees-of-freedom reduced-order lumped parameter mass model is derived by neglecting high-frequency vibration modes. The neglected vibration modes constitute the unstructured system uncertainties. An attempt has been made to reduce the unmodeled uncertainties by placing actuators and/or sensors at the node points of a neglected mode. The H _∞-based control law is able to suppress the low-order vibration modes without any spillover instability due to neglected modes. The proposed control scheme is also shown to be robust against parameter variations. The performance of the control scheme is verified both by simulation and experimental studies     

A fuzzy sliding controller for nonlinear systems

It is well known that sliding-mode control can give good transient performance and system robustness. However, the presence of chattering may introduce problems to the actuators. Many chattering elimination methods use a finite DC gain controller which leads to a finite steady-state error. One method to ensure zero steady-state error is using a proportional plus integral (PI) controller. This paper proposes a fuzzy logic controller which combines a sliding-mode controller (SMC) and a PI controller. The advantages of the SMC and the PI controller can be combined and their disadvantages can be removed. The system stability is proved, although there is one more state variable to be considered in the PI subsystem. An illustrative example shows that good transient and steady-state responses can be obtained by applying the proposed controller