A novel data based control method based upon neural network and simultaneous perturbation stochastic approximation

A novel data based control method is proposed by modifying the cost function of a model-free control method based on Simultaneous Perturbation Stochastic Approximation (SPSA). The controller is constructed through use of a Function Approximator (FA), which is fixed as a neural network here. In the novel approach, the ability of the controller has been greatly improved. At last, the proposed modified control method is applied to solve non-linear tracking problems. Simulation comparison tests were done on typical non-linear plants, through which, the effectiveness of the novel data based control method is fully illustrated.     

Real-time robust adaptive control of robots subjected to actuator voltage constraint

This paper is concerned with the problem of design and implementation of a robust adaptive control strategy for electrically driven robots while considering to the constraints on the actuator voltage input. The proposed approach provides a flexible design framework and stable to deal with robustness compared with many other adaptive controllers, such as halting/slowing adaption techniques and adaptively adjusting command signal, which are proposed for robotic applications. The control design procedure is based on a new form of universal approximation theory and using Stone–Weierstrass theorem, to avoid saturation besides being robust against both structured and unstructured uncertainties associated with external disturbances and actuated manipulator dynamics. Moreover, the proposed approach eliminates problems arising from classic adaptive feedforward control scheme. The analytical studies as well as experimental results produced using MATLAB/SIMULINK external mode control on a two degree of freedom electrically driven robot demonstrate high performance of the proposed control schemes.     

Robust control of electrically driven robots by adaptive fuzzy estimation of uncertainty

This paper presents a novel robust decentralized control of electrically driven robot manipulators by adaptive fuzzy estimation and compensation of uncertainty. The proposed control employs voltage control strategy, which is simpler and more efficient than the conventional strategy, the so-called torque control strategy, due to being free from manipulator dynamics. It is verified that the proposed adaptive fuzzy system can model the uncertainty as a nonlinear function of the joint position error and its time derivative. The adaptive fuzzy system has an advantage that does not employ all system states to estimate the uncertainty. The stability analysis, performance evaluation, and simulation results are presented to verify the effectiveness of the method. A?comparison between the proposed Nonlinear Adaptive Fuzzy Control (NAFC) and a Robust Nonlinear Control (RNC) is presented. Both control approaches are robust with a very good tracking performance. The NAFC is superior to the RNC in the face of smooth uncertainty. In contrast, the RNC is superior to the NAFC in the face of sudden changes in uncertainty. The case study is an articulated manipulator driven by permanent magnet dc motors.     

Decentralized direct adaptive fuzzy control of robots using voltage control strategy

Decentralized control is the most favorite control of robot manipulators due to computational simplicity and ease of implementation. Beside that, adaptive fuzzy control efficiently controls uncertain nonlinear systems. These motivate us to design a decentralized fuzzy controller. However, there are some challenging problems to guarantee stability. The state-space model of the robotic system including the robot manipulator and motors is in a noncompanion form, multivariable, highly nonlinear, and heavily coupled with a variable input gain matrix. For this purpose, adaptive fuzzy control may use all variable states. As a result, it suffers from computational burden. To overcome the problems, we present a novel decentralized Direct Adaptive Fuzzy Control (DAFC) of electrically driven robot manipulators using the voltage control strategy. The proposed DAFC is simple, in a decentralized structure with high-accuracy response, robust tracking performance, and guaranteed stability. Instead of all state variables, only the tracking error of every joint and its derivative are given as the inputs of the controller. The proposed DAFC is simulated on a SCARA robot driven by permanent magnet dc motors. Simulation results verify superiority of the decentralized DAFC to a decentralized PD-fuzzy controller.     

猎雷声纳基阵姿态稳定系统自适应反步法控制

针对带非线性摩擦力矩和负载扰动的高精度猎雷声纳基阵姿态稳定系统,提出了一种基于神经网络的自适应反步法控制方法。其中神经网络用于估计未知非线性摩擦力矩,进而设计反步法控制器和参数自适应律来对神经网络估计误差和负载扰动进行补偿。最后应用Lyapunov方法证明了所提出的自适应控制器能保证闭环系统的稳定性,并且可以通过选择适当的控制器参数来调整收敛率。仿真结果表明,基于神经网络的自适应反步法控制方法与PID控制相比,系统的动、静态性能指标及鲁棒性得到了全面的改善,与双闭环PID控制相比,跟踪精度提高了3倍多。     

Finite-time synchronization control for a class of perturbed nonlinear systems with fixed convergence time and hysteresis quantizer: applied to Genesio–Tesi chaotic system

In the present article, a terminal sliding mode control strategy has been proposed in order to address the synchronization problem for a class of perturbed nonlinear systems with fixed convergence time and input quantization. The proposed protocol guarantees the fixed-time convergence of the sliding manifold to the origin, which means that the convergence time of the proposed sliding manifold does not change on the variations of initial values, different from typical control methods. Here, the hysteresis quantizer, as a specific type of quantizer with nonlinear sector-bounded, is applied in order to quantize the input signal. The proposed quantized control scheme vigorously prevents the potential adverse chattering phenomenon which is experienced in the common quantization methods. The proposed controller does not need the limiting criteria related to considered parameters of quantization compared to recent control approaches. Finally, the designed controller is implemented on the perturbed Genesio–Tesi (G–T) chaotic systems to verify effectiveness and strength of the proposed method.

     

A novel ADP based model-free predictive control

Dynamic programming is a very useful tool in solving optimization and optimal control problems. Here, the Approximate Dynamic Programming (ADP) and the notion of neural networks based predictive control are combined with a model-free control method based on SPSA (Simultaneous perturbation stochastic approximation), and a novel ADP based model-free predictive control strategy for nonlinear systems is proposed. Dynamic programming is used to adjust the control parameters in the novel model-free control method and the notion of predictive control is introduced to modify the whole control structure. Finally, the proposed ADP based model-free predictive control strategy is applied to solve nonlinear tracking problems and the effectiveness of this novel control method is fully illustrated though simulation tests on two typical nonlinear systems.     

Adaptive interval type-2 fuzzy sliding mode control for unknown chaotic system

In this paper, a novel adaptive interval type-2 fuzzy sliding mode control (AIT2FSMC) methodology is proposed based on the integration of sliding mode control and adaptive interval type-2 fuzzy control for chaotic system. The AIT2FSMC system is comprised of a fuzzy control design and a hitting control design. In the fuzzy control design, an interval type-2 fuzzy controller is designed to mimic a feedback linearization (FL) control law. In the hitting control design, a hitting controller is designed to compensate the approximation error between the FL control law and the interval type-2 fuzzy controller. The parameters of the interval type-2 fuzzy controller, as well as the uncertainty bound of the approximation error, are tuned adaptively. The adaptive laws are derived in the sense of Lyapunov stability theorem, thus the stability of the system can be guaranteed. The proposed control system compared to adaptive fuzzy sliding mode control (AFSMC). Simulation results show that the proposed control systems can achieve favorable performance and robust with respect to system uncertainties and external disturbances.     

Global convergence of inexact Newton methods for transonic flow

In computational fluid dynamics, non-linear differential equations are essential to represent important effects such as shock waves in transonic flow. Discretized versions of these non-linear equations are solved using iterative methods. In this paper an inexact Newton method using the GMRES algorithm of Saad and Schultz is examined in the context of the full potential equation of aerodynamics. In this setting, reliable and efficient convergence of Newton methods is difficult to achieve. A poor initial solution guess often leads to divergence or very slow convergence. This paper examines several possible solutions to these problems, including a standard local damping strategy for Newton's method and two continuation methods, one of which utilizes interpolation from a coarse grid solution to obtain the initial guess on a finer grid. It is shown that the continuation methods can be used to augment the local damping strategy to achieve convergence for difficult transonic flow problems. These include simple wings with shock waves as well as problems involving engine power effects. These latter cases are modelled using the assumption that each exhaust plume is isentropic but has a different total pressure and/or temperature than the freestream.     

Sequential non-linear least-square estimation for damage identification of structures

An early detection of structural damage is an important goal of any structural health monitoring system. In particular, the ability to detect damages on-line, based on vibration data measured from sensors, will ensure the reliability and safety of the structures. In this connection, innovative data analysis techniques for the on-line damage detection of structures have received considerable attentions recently, although the problem is quite challenging. In this paper, we proposed a new data analysis method, referred to as the sequential non-linear least-square (SNLSE) approach, for the on-line identification of structural parameters. This new approach has significant advantages over the extended Kalman filter (EKF) approach in terms of the stability and convergence of the solution as well as the computational efforts involved. Further, an adaptive tracking technique recently proposed has been implemented in the proposed SNLSE to identify the time-varying system parameters of the structure. The accuracy and effectiveness of the proposed approach have been demonstrated using the Phase I ASCE structural health monitoring benchmark building, a non-linear elastic structure and non-linear hysteretic structures. Simulation results indicate that the proposed approach is capable of tracking on-line the changes of structural parameters leading to the identification of structural damages.     

Adaptive terminal sliding mode control for nonlinear differential inclusion systems with disturbance

This paper deals with the adaptive terminal sliding mode control for nonlinear differential inclusion systems subjected to disturbance. The upper bound of the disturbance is unknown. First, the fast terminal sliding mode surface is established and sufficient condition for fast convergence is given. Then the adaptive sliding mode controller is designed to make the state of system arrive at the sliding mode in finite time. A numerical example is provided to show the effectiveness of the proposed method.     

Development of adaptive modeling techniques for non-linear hysteretic systems

Adaptive estimation procedures have gained significant attention by the research community to perform real-time identification of non-linear hysteretic structural systems under arbitrary dynamic excitations. Such techniques promise to provide real-time, robust tracking of system response as well as the ability to track time variation within the system being modeled. An overview of some of the authors’ previous work in this area is presented, along with a discussion of some of the emerging issues being tackled with regard to this class of problems. The trade-offs between parametric-based modeling and non-parametric modeling of non-linear hysteretic dynamic system behavior are discussed. Particular attention is given to (1) the effects of over- and under-parameterization on parameter convergence and system output tracking performance, (2) identifiability in multi-degree-of-freedom structural systems, (3) trade-offs in setting user-defined parameters for adaptive laws, and (4) the effects of noise on measurement integration. Both simulation and experimental results indicating the performance of the parametric and non-parametric methods are presented and their implications are discussed in the context of adaptive structures and structural health monitoring.     

STOCHASTIC OPTIMAL CONTROL OF STRONGLY NONLINEAR SYSTEMS UNDER WIDE-BAND RANDOM EXCITATION WITH ACTUATOR SATURATION

A bounded optimal control strategy for strongly non-linear systems under non-white wide-band random excitation with actuator saturation is proposed. First, the stochastic averaging method is introduced for controlled strongly non-linear systems under wide-band random excitation using generalized harmonic functions. Then, the dynamical programming equation for the saturated control problem is formulated from the partially averaged Itō equation based on the dynamical programming principle. The optimal control consisting of the unbounded optimal control and the bounded bang-bang control is determined by solving the dynamical programming equation. Finally, the response of the optimally controlled system is predicted by solving the reduced Fokker-Planck-Kolmogorov （FPK） equation associated with the completed averaged Itō equation. An example is given to illustrate the proposed control strategy. Numerical results show that the proposed control strategy has high control effectiveness and efficiency and the chattering is reduced significantly comparing with the bang-bang control strategy.     

Modeling of a magnetorheological damper by recursive lazy learning

Nowadays dampers based on magnetorheological (MR) fluids are receiving significant attention specially for control of structural vibration and automotive suspensions systems. In most cases, it is necessary to develop an appropriate control strategy which is practically implementable when a suitable model for MR dampers is available. It is not a trivial task to model the dynamic of MR dampers because of their inherent non-linear and hysteretic dynamics. In this paper, a recursive lazy learning method based on neural networks is considered to model the MR damper behavior. The proposed method is validated by comparison with experimental obtained responses. Results show the estimated model correlates very well with the data obtained experimentally. The method proposed learns quickly that it is only necessarily a learning cycle, it can learn on-line and it is easy to select the network structure and calculate the model parameters.     

Non-linear cancellation of the parametric resonance in elastic beams: Theory and experiment

A non-linear control strategy is applied to a simply supported uniform elastic beam subjected to an axial end force at the principal-parametric resonance frequency of the first skew-symmetric mode. The control input consists of the bending couples applied by two pairs of piezoceramic actuators attached onto both sides of the beam surfaces and symmetrically with respect to the midspan, driven by the same voltage, thus resulting into symmetric control forces. This control architecture has zero control authority, in a linear sense, onto skew-symmetric vibrations. The non-linear transfer of energy from symmetric motions to skew-symmetric modes, due to non-linear inertia and curvature effects, provides the key physical mechanism for channelling suitable control power from the actuators into the linearly uncontrollable mode. The reduced dynamics of the system, constructed with the method of multiple scales directly applied to the governing PDE’s and boundary conditions, suggest effective forms of the control law as a two-frequency input in sub-combination resonance with the parametrically driven mode. The performances of different control laws are investigated. The relative phase and frequency relationships are designed so as to render the control action the most effective. The control schemes generate non-linear controller forces which increase the threshold for the activation of the parametric resonance thus resulting into its annihilation. The theoretical predictions are compared with experimentally obtained results.     

Direct adaptive neural control of chaos in the permanent magnet synchronous motor

In this paper, a direct adaptive neural speed tracking control is addressed for the chaotic permanent magnet synchronous motor (PMSM) drive systems via backstepping. Neural networks are directly used to approximate unknown and desired control signals and a novel direct adaptive tracking controller is constructed via backstepping. The proposed adaptive neural controllers guarantee that the tracking error converges to a small neighborhood of the origin. Compared with the conventional backstepping method, the designed neural controller??s structure is very simple. Simulation results show that the proposed control scheme can suppress the chaos of PMSM and guarantees the perfect tracking performance even with the existence of unknown parameters.     

Nonlinear control of electrical flexible-joint robots

This paper is devoted to the nonlinear tracking control of electrically driven flexible-joint manipulators using the voltage control strategy. Despite the torque control laws that are involved in the complexity of manipulator dynamics, the proposed control law is free from manipulator dynamics. This novelty is for adopting the voltage control strategy to derive a simple robust adaptive control under both structured and unstructured uncertainty. The proposed control approach has a fast response with a good tracking performance under the well-behaved control efforts in the form of decentralized control. The control method is justified by the stability analysis and simulated on a flexible-joint electrically driven robot manipulator.     

Decentralized adaptive attitude synchronization control for spacecraft formation using nonsingular fast terminal sliding mode

This paper studies the attitude synchronization control problem for a group of spacecraft. Considering inertia uncertainties and external disturbances with unknown bounds, a decentralized adaptive control scheme is developed using nonsingular fast terminal sliding mode (NFTSM). A multispacecraft NFTSM is firstly designed, which contains the advantages of the nonsingular terminal sliding mode and the traditional linear sliding mode together. Then, the continuous decentralized adaptive NFTSM control laws with boundary layer by employing NFTSM associated with novel adaptive architecture are proposed, which can eliminate the chattering, and guarantee the attitude tracking errors converge to the regions containing the origin in finite time. At last, numerical simulations are presented to demonstrate the performance of the proposed control strategy.     

Robust adaptive nonlinear control for uncertain control-affine systems and its applications

This paper addresses the robust tracking control problem for a class of uncertain nonlinear systems with time-varying parameters, perturbed by external disturbances. The unknown time-varying parameters and disturbances are neither required to be periodic nor to have known bounds. Depending on the characteristics of disturbance signals, two adaptive-based control algorithms are developed. First, an adaptive H _∞ control is designed that achieves: (i) an H _∞ tracking performance when the external disturbances are L ₂ signals, and (ii) the convergence of tracking error to zero if the disturbances are bounded and L ₂ signals. Then a novel adaptive control algorithm is proposed, only with the assumption of boundedness of disturbances, to drive the tracking error to zero. The designed tracking controllers are then used for controlling a cart-pendulum system, as an underactuated mechanical system, and chaos synchronization of uncertain Genesio–Tesi chaotic system. Numerical simulations are also given to demonstrate the effectiveness of the proposed control schemes.     

Distributed adaptive tracking control of multiple flexible spacecraft under various actuator and measurement limitations

This paper investigates the distributed leader–follower tracking problem for a team of flexible spacecraft over an undirected communication network with uncertain parameters subject to various actuator and measurement limitations. Assuming that at least one team member can receive information from the virtual leader, three scenarios are considered: (i) all the states of the flexible spacecraft can be completely measured and driven, (ii) only the rigid part of the flexible spacecraft can be driven with full state feedback and (iii) only the rigid part of the flexible spacecraft can be measured and driven. In the first case, a continuous adaptive control law is designed by building a unified architecture based on the linear-in-parameter property. In the second case, a distributed adaptive control algorithm is developed with a discontinuous parameter update law by treating the team of flexible spacecraft as two cascading subsystems. In the third case, a distributed adaptive control law is established with feedback from the generalized coordinates, generalized velocities and generalized accelerations of the rigid part of the spacecraft. It is theoretically proved that the closed-loop systems under the three designed adaptive control laws are all convergent to the target states. Finally, three numerical examples are presented to illustrate the effectiveness of the three proposed control laws.