摘要:

A fast and concise MTMO nonlinear model validity test procedure is derived, based on higher order correlation functions, to form a global-to-local hierarchical validation diagnosis of identified MEMO linear and nonlinear models. The new procedure is applied to four MIMO nonlinear system models including a neural network training example, to demonstrate the effectiveness of the tests.

ISSN:0020-7179    

DOI:10.1080/00207179508921566

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

A study is made of a simple but realistic control problem that does not satisfy one of the conditions for polynomial optimization. The use of both state-space/Kalman filter and polynomial methods illustrates the advantage of the latter with regard to simplicity and the detection of flaws. Two solutions to the problem are proposed.

ISSN:0020-7179    

DOI:10.1080/00207179508921567

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

This paper considers the problem of decentralized fuzzy control of multivariable systems. Some definitions and theorems with regard to this problem are given. Decentralized fuzzy control algorithms, based on active decomposition of control laws, are presented and illustrated by numerical examples. The algorithms use local sets of fuzzy relations, decoupling the influence of state on control variables when added to the original fuzzy relations. It is shown that the number of fuzzy relations is significantly reduced and thus real-time control implementation is facilitated.

ISSN:0020-7179    

DOI:10.1080/00207179508921568

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

A new digital controller named aβ–predictive regulator is proposed. The digital control system design procedure is extremely simple. The discrete process model for the sampling periodβδt,β > 1, for the control sampling period δtis determined. This model is applied at timet = kδtas the one-step ahead predictor for the controlled output. Equalizing the predicted controlled outputyp((k + 1)δt) with the desired process outputyr((k + 1)δt)obtains the control inputu(k) The desired process outputyr((k + 1)δt) is defined, as in model algorithmic control, by a first-order reference trajectory, with the speed of response defined by a parameter |αβ| < 1. This parameter is used for direct adjustment of the closed-loop speed of response. The parameterβis used for stabilization of the closed-loop system. The performance of the design method proposed is demonstrated by some simulation results.

ISSN:0020-7179    

DOI:10.1080/00207179508921569

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

Using mappings of the form (1 + bf)a, (1 − bf)−aand ((1 + bf)/(1 − bf))a, we get three different unit interpolation algorithms, respectively first, second and third. The second algorithm yields a unit in H ∞ with arbitrary specified left half j-plane zeros instead of arbitrary specified left half s-plane poles, as in Youla et al. (1974) and Vidyasagar (1985) with the first algorithm. The third algorithm yields a unit in H ∞ with considerably lower degree in the cases where a is required to be greater than one with either the first or second algorithm. However, by using alternative steps from the first and second algorithms judiciously, as shown by a numerical example, one can reduce the order of the unit considerably, compared with the third algorithm. The usefulness of the second algorithm in the control system context is that the closed-loop system poles can be specified arbitrarily in the LHP in a strong stabilization problem, which is the opposite to the first algorithm of Youlaet al. (1974) and Vidyasagar (1985), where the controller poles can be specified in the LHP in an arbitrary way. The second algorithm can also be used for placing the interconnected closed loop system poles in decentralized stabilization for expanding construction of large-scale systems.

ISSN:0020-7179    

DOI:10.1080/00207179508921570

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

We consider the stabilization of a class of multivariable nonlinear systems, about an equilibrium point at the origin, using variable structure output feedback control. In particular, the system can be transformed into a normal form with no zero dynamics. A robust high-grain observer is used to estimate the state variables while rejecting the effect of disturbance. A globally bounded discontinuous variable structure controller is designed to compensate for modelling error. We show that the controller can stabilize the closed-loop system and does not suffer from the peaking phenomenon that exists in previous designs.

ISSN:0020-7179    

DOI:10.1080/00207179508921571

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

We present a control law for globally asymptotically stabilizing a class of controllable nonlinear systems without drift. The control law converts into closed loop feedback earlier strategies for open loop steering of non-holonomic systems using sinusoids at integrally related frequencies. The global result is obtained by introducing saturation functions. Simulation results for stabilizing a simple kinematic model of an automobile are included.

ISSN:0020-7179    

DOI:10.1080/00207179508921572

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

The design problem of robust stabilizing controllers for nonlinear uncertain systems is considered in this paper. The nominal system is input-output linearizable, and the uncertainty is assumed to be either generalized matching or mismatched. The control design combines differential geometric feedback linearization with the deterministic approach, in which no statistical property of the uncertainty is assumed and used; the only information on the uncertainty, which is needed for the design, is its possible bound. Two robust stabilizing controllers for generalized matching uncertain systems are given. For mismatched uncertain systems, a systematic design procedure is first presented, then four robust stabilizing controllers are given. The robust controllers, which are given in forms of explicit feedback laws based on the nominal system and are parametrized by the robust design parameters related to the bounds of uncertainties, are more practicable than the existing ones in the literature. Numerical examples and simulation results verify the effectiveness of the design method and controllers.

ISSN:0020-7179    

DOI:10.1080/00207179508921573

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

A principle for the design of control systems, called matching, requires primarily that the set of all possible inputs to a system is a subset of the set of all tolerable inputs (those which are judged to produce an acceptable response). In this paper, a readily computable necessary and sufficient condition for matching is provided for the case in which the tolerable set is defined by uniform bounds on the magnitude of outputs, and the possible set is defined by separate restrictions on the integral square of the input and the integral square of the input's rate of change. An example of a matched design obtained using the condition is given.

ISSN:0020-7179    

DOI:10.1080/00207179508921574

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

The problem of output tracking in SISO uncertain non-autonomous nonlinear systems is treated in this study. It is shown that a nonlinear system experiencing non-matching disturbances and uncertainties can be transformed to a form similar to the normal form. Then, a novel variable structure controller with disturbance estimation and parameter adaptation is presented. This controller, while preserving the robustness properties of sliding control, avoids a sliding mode and generates a continuous control signal, which results in close output tracking.

ISSN:0020-7179    

DOI:10.1080/00207179508921575

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor