J. Q. Gong

Modeling and cancellation of pivot nonlinearity in hard disk drives

This paper considers the issue of modeling the pivot nonlinearity in a typical hard disk drive, taking into account viscous friction, Coulomb friction and hysteresis effects. It presents a nonlinear compensator design, based on a proposed model, to make the compensated system behave like a double integrator regardless of the magnitude of the input current. It describes extensive experiments, carried out in both time domain and frequency domain, to show the effects of pivot nonlinearity and the effectiveness of the compensator.

Adaptive robust control without knowing bounds of parameter variations

A discontinuous projection based adaptive robust control (ARC) design is constructed for nonlinear systems transformable to the semi-strict feedback form without knowing the bounds of parameter variations and uncertain nonlinearities. The proposed ARC design only adapts actual physical parameters and uses a fixed design bound plus certain robust feedback to account for the possible destabilizing effect of online parameter adaptation. The resulting ARC controller achieves a prescribed transient performance and final tracking accuracy in general; the exponential convergence rate of the transient tracking error and the bound of the final tracking error can be adjusted via certain controller parameters in a known form. In addition, in the presence of parametric uncertainties only, if the true parameters fall within the design bound, asymptotic output tracking is achieved. Furthermore, by choosing the design bound appropriately, the controller also has a well-designed built-in anti-integration windup mechanism to alleviate the effect of control saturation.

Modeling and cancellation of pivot nonlinearity in hard disk drive

In this paper, the modeling issue of the pivot nonlinearity in a typical hard disk drive is considered. Viscous friction, Coulomb friction and hysteresis effects are taken into account. Based on the proposed model, a nonlinear compensator is designed to make the compensated system to behave like a double integrator regardless of the magnitude of the input current. Extensive experiments are then carried out in both time domain and frequency domain to show the effects of pivot nonlinearity and the effectiveness of the proposed compensator.

Neural network adaptive robust control of nonlinear systems in semi-strict feedback form

In this paper, the recently proposed neural network adaptive robust control (NNARC) design axe generalized to synthesize performance oriented control laws for a class of nonlinear systems transformable to the semi-strict feedback forms through the incorporation of backstepping design techniques. All unknown but repeatable nonlinearities in system are approximated by outputs of multi-layer neural networks to achieve a better model compensation and an improved performance. Through the use of discontinuous projections with fictitious bounds, a controlled on-line training of all NN weights is achieved. Robust control terms can then be constructed to attenuate various model uncertainties effectively for a guaranteed output tracking transient performance and a guaranteed final tracking accuracy.

Fuzzy adaptive robust control of a class of nonlinear systems

A fuzzy adaptive robust tracking controller for a class of uncertain nonlinear dynamical systems is proposed and analyzed. The controllers construction and its analysis involve sliding modes. The proposed controller consists of two components. A robust feedback component is employed to eliminate the effects of disturbances, while a fuzzy logic component equipped with an adaptation mechanism reduces modeling uncertainties by approximating the models nonlinearities on-line. A projection method is used to prevent the adaptation parameters from going unbounded in the presence of disturbances. It is shown that the closed-loop system driven by the proposed controller is stable and the adaptation parameters are bounded. A guaranteed transient performance and a guaranteed final tracking accuracy in the presence of parametric uncertainties and disturbances are achieved. Furthermore, if there are no disturbances and the unknown models nonlinearities are within the approximation range of the fuzzy logic system, asymptotic output tracking is also achieved.

Neural network-based adaptive robust control of a class of nonlinear systems in normal form

Neural networks (NNs) and adaptive robust control (ARC) design philosophy are integrated to design performance oriented control laws for a class of n-th order nonlinear systems in a normal form in the presence of both repeatable and non-repeatable uncertain nonlinearities. Unknown nonlinearities can exist in the input channel also. All unknown but repeatable nonlinearities are approximated by outputs of multi-layer NNs. A discontinuous projection method with fictitious bounds is used to tune NN weights online with no prior information for a controlled learning process. Robust terms are constructed to attenuate model uncertainties effectively for a guaranteed output tracking transient performance and a guaranteed final tracking accuracy. If the unknown nonlinear functions are in the functional ranges of NNs and the ideal weights fall within the prescribed range, asymptotic output backing is also achieved. Furthermore, by choosing the prescribed range appropriately, the controller may have a well-designed built-in anti-integration windup mechanism.

Output feedback neural network adaptive robust control with application to linear motor drive system

In this paper, neural networks (NNs) and adaptive robust control design method are integrated to design a performance oriented control law with only output feedback for a class of single-input-single-output nth order nonlinear systems in a normal form. The nonlinearities in the system include repeatable unknown nonlinearities and nonrepeatable unknown nonlinearities such as external disturbances. In addition, unknown nonlin-earities can exist in the control input channel as well. A high-gain observer is employed to estimate the states of system. All unknown but repeatable nonlinear function are approximated by the outputs of multilayer neural networks with the estimated states as inputs to achieve a better model compensation. All NN weights are tuned on-line. In order to avoid possible divergence of on-line tuning, discontinuous projections with fictitious bounds are used in the weight adjusting law to make sure that all the weight are adapted within a prescribed range. Theoretically, the resulting controller achieves a guaranteed output tracking transient performance and a guaranteed final tracking accu-racy in general. Certain robust control terms is then constructed to effectively attenuate various model uncertainties and estimate errors. Furthermore, if all the states are available and the unknown nonlinear functions are in the functional ranges of the neural networks, an asymptotic output tracking is also achieved to retain the perfect learning capability of NNs in the ideal situation provided that the ideal NN weights fall within the prescribed range. The output feedback neural network adaptive robust control is then applied to the control of a linear motor drive system. Experiments are carried out to show the effectiveness of the proposed algorithm and the excellent output tracking performance.

Output feedback neural network adaptive robust control of a class of SISO nonlinear systems

Through the use of high-gain observer to estimate the unmeasurable system states, neural networks (NN) and adaptive robust control (ARC) method are integrated to design performance oriented output feedback control laws for a class of single-input-single-output (SISO) nth-order nonlinear systems in normal form. Multi-layer neural networks (MLNN) with the estimated states as inputs are used to approximate all unknown but repeatable nonlinear functions in the system. A controlled learning is achieved through the use of discontinuous projections with fictitious bounds in the tuning laws for NN weights. Certain robust control terms are constructed to effectively attenuate various model uncertainties and estimation errors for a guaranteed output tracking transient performance and a guaranteed final tracking accuracy in general. In addition, asymptotic output tracking is achieved in the ideal case. Experimental results on the output feedback control of a linear motor drive system are obtained to illustrate the proposed algorithm.

Indirect Neural Network Adaptive Robust Control of Linear Motor Drive System

In this paper, an indirect neural network adaptive robust control (INNARC) scheme is developed for the precision motion control of linear motor drive systems. The proposed INNARC achieves not only good output tracking performance but also excellent identifications of unknown nonlinear forces in system for secondary purposes such as prognostics and machine health monitoring. Such dual objectives are accomplished through the complete separation of unknown nonlinearity estimation via neural networks and the design of baseline adaptive robust control (ARC) law for output tracking performance. Specifically, recurrent neural network (NN) structure with NN weights tuned on-line is employed to approximate various unknown nonlinear forces of the system having unknown forms to adapt to various operating conditions. The design is actual system dynamics based, which makes the resulting on-line weight tuning law much more robust and accurate than those in the tracking error dynamics based direct NNARC designs in implementation. With a controlled learning process achieved through projection type weights adaptation laws, certain robust control terms are constructed to attenuate the effect of possibly large transient modelling error for a theoretically guaranteed robust output tracking performance in general. Experimental results are obtained to verify the effectiveness of the proposed INNARC strategy. For example, for a typical point-to-point movement, with a measurement resolution level of ±1μm, the output tracking error during the entire execution period is within ±5μm and mainly stays within ±2μm showing excellent output tracking performance. At the same time, the outputs of NNs approximate the unknown forces very well allowing the estimates to be used for secondary purposes such as prognostics.Copyright