Search

We consider a Thévenin equivalent circuit capturing the dynamics of a power grid as seen from the point of common coupling with a power electronic converter and propose a solution to the problem of online identification of the corresponding circuit parameters. For this purpose, we first derive a linear regression model in the conventional abc coordinates that, unfortunately, suffers from severe lack of identifiability properties. In spite of this fact, we design a bounded observer‐based composite identifier that, requiring only local measurements and knowledge of the grid frequency, ensures global boundedness of the parameters estimation error and convergence to a practically small residual set. An extension that guarantees exponential convergence of these errors to zero, in the case that the grid is dominantly inductive, is further provided. The performance of the proposed identifier, which subsumes a conventional gradient descent algorithm, is illustrated via detailed computer simulations. [ABSTRACT FROM AUTHOR]

In this article, we are interested in the problem of adaptive synchronization of a voltage source converter with a possibly weak grid with unknown angle and frequency. To guarantee a suitable synchronization with the angle of the three‐phase grid voltage we design an adaptive observer for such a signal requiring measurements only at the point of common coupling. Then we propose an alternative certainty‐equivalent, adaptive phase‐locked loop that ensures the angle estimation error goes to zero for almost all initial conditions. Although well‐known, for the sake of completeness, we also present a PI controller with feedforward action that ensures the converter currents converge to an arbitrary desired value. Relevance of the theoretical results and their robustness to variation of the grid parameters are thoroughly discussed and validated in the challenging scenario of a converter connected to a grid with low short‐circuit‐ratio. [ABSTRACT FROM AUTHOR]

In this article, we address the problem of adaptive state observation of linear time‐varying systems with delayed measurements and unknown parameters. Our new developments extend the results reported in our recently works. The case with known parameters has been studied by many researchers. However in this article we show that the generalized parameter estimation‐based observer design provides a very simple solution for the unknown parameter case. Moreover, when this observer design technique is combined with the dynamic regressor extension and mixing estimation procedure the estimated state and parameters converge in fixed‐time imposing extremely weak excitation assumptions. [ABSTRACT FROM AUTHOR]

Summary: In this paper we address the problem of distributed state estimation of continuous‐ and discrete‐time stable, LTI systems. The classical observer canonical form representation of the system dynamics is used to identify the observable states for each node (agent). The main novelty of our contribution is to obviate the need of the asymptotically (or finite‐time) local Luenberger observers. Instead, following the generalized parameter estimation‐based observer design approach, we use the state transition matrix of the system to translate the problem of reconstruction of the observable states into one of parameter estimation‐‐‐namely, the initial conditions of the associated trajectory. Exploiting the local observability property of the agents we prove that, with a simple sample‐and‐hold or finite summation operation, it is possible to estimate these parameters algebraically. Consensus strategies are then used to fuse the parameter estimates of all agents to reconstruct the complete state vector. Asymptotic or finite convergence time of the observer are established for several scenarios for the graph, including time‐varying, switching and with transmission delays. [ABSTRACT FROM AUTHOR]

Summary: In this article, we propose a PI passivity‐based controller, applicable to a large class of switched power converters, that ensures global state regulation to a desired equilibrium point. A solution to this problem requires full state‐feedback, which makes it practically unfeasible. To overcome this limitation we construct a state observer that is implementable with measurements that are available in practical applications. The observer reconstructs the state in finite‐time, ensuring global convergence of the PI. An adaptive version of the observer, where some parameters of the converter are estimated, is also proposed. The excitation requirement for the observer is very weak and is satisfied in normal operation of the converters. Realistic simulation results illustrate the excellent performance and robustness vis‐à‐vis noise and parameter uncertainty of the proposed output‐feedback PI. [ABSTRACT FROM AUTHOR]

Summary: In this article, we address the problems of flux and speed observer design for voltage‐fed induction motors with unknown rotor resistance and load torque. The only measured signals are stator current and control voltage. Invoking the recently reported Dynamic Regressor Extension and Mixing‐Based Adaptive Observer (DREMBAO), we provide the first global solution to this problem. The proposed DREMBAO achieves asymptotic convergence under an excitation condition that is strictly weaker than persistent excitation. If the latter condition is assumed the convergence is exponential. [ABSTRACT FROM AUTHOR]

This paper is devoted to the application of a recently proposed globally convergent adaptive position observer to non‐salient permanent magnet synchronous motors. Following the Dynamic Regressor Extension and Mixing Based Adaptive Observer (DREMBAO) approach, a new finite‐time robust observer is presented that allows to track adaptively the rotor position by measuring only the currents and voltages and without knowledge of mechanical, electrical and magnetic parameters. A numerical example for the case with different rotor speeds and time‐varying load torque is considered to reveal the advantages of the proposed approach in comparison with other existing methods. Experimental studies of the proposed robust nonlinear observer implementation are presented to illustrate the efficiency of the new design technique in different speed modes together with adaptive estimation of unknown parameters. [ABSTRACT FROM AUTHOR]

The authors propose a novel decentralised mixed algebraic and dynamic state observation method for multi-machine power systems with unknown inputs and equipped with phasor measurement units (PMUs). More specifically, they prove that for the third-order flux-decay model of a synchronous generator, the local PMU measurements provide enough information to reconstruct algebraically the load angle and quadrature-axis internal voltage. Due to the algebraic structure, a high numerical efficiency is achieved, which makes the method applicable to large-scale power systems. Also, they prove that the relative shaft speed can be globally estimated combining a classical immersion and invariance observer with – the recently introduced – dynamic regressor extension and mixing parameter estimator. This adaptive observer ensures global convergence under weak excitation assumptions that are verified in applications. The proposed method neither requires the measurement of exogenous input signals such as the field voltage and the mechanical torque nor the knowledge of mechanical subsystem parameters. [ABSTRACT FROM AUTHOR]

One of the most challenging problems in AC drives applications is the design of a simple plug‐in adaptation scheme to estimate the unknown rotor resistance and load torque for the industry‐standard indirect field oriented control. In this paper we give the first globally convergent solution to this problem for torque control of current‐fed induction motors that does not rely on any excitation assumption. Some results on speed regulation are also presented. [ABSTRACT FROM AUTHOR]

Summary: This paper considers the problem of continuous‐time online frequency estimation for a biased damped sinusoidal signal. The previous result for a sinusoidal signal with time‐varying amplitude requires a persistency of excitation condition for regressor, which is not satisfied in the considered case. To relax this condition, we propose to use Dynamic Regressor Extension and Mixing method on the first step to replace nth‐order regression with n first‐order regression models. On the second step, two simple relaxation methods are proposed to establish necessary excitation for the first‐order gradient‐based estimator. The efficiency of the proposed approach is demonstrated through the set of numerical simulations for the exponentially damped sinusoidal signal. [ABSTRACT FROM AUTHOR]

Summary: In this paper, we propose the first solution to the long‐standing problem of designing a globally convergent direct adaptive pole‐placement controller for linear, time‐invariant discrete‐time systems with arbitrary zeros that does not rely on persistency of excitation assumptions. As is well known, the main difficulty of this design is that it involves the estimation of parameters that enter nonlinearly in the regression model. This problem can be overcome introducing an overparameterized representation of the system, which imposes very strict persistency of excitation conditions to prove the parameter convergence. The latter is avoided here using a new version of the dynamic regressor extension and mixing parameter estimator recently proposed in the literature. The main feature of this estimator is that it generates, out of an m‐dimensional vector regression, m scalar regression models. This property allows us to estimate only the controller parameters of interest for the adaptive implementation, whose convergence is ensured under assumptions that are strictly weaker than the classical persistency of excitation requirement. Simulation results that illustrate the performance of the proposed scheme are also presented. [ABSTRACT FROM AUTHOR]

Summary: Parameter convergence is desirable in adaptive control as it enhances the overall stability and robustness properties of the closed‐loop system. In existing online historical data (OHD)–driven parameter learning schemes, all OHD are exploited to update parameter estimates such that parameter convergence is guaranteed under a sufficient excitation (SE) condition which is strictly weaker than the classical persistent excitation condition. Nevertheless, the exploitation of all OHD not only results in possible unbounded adaptation but also loses the flexibility of handling slowly time‐varying uncertainties. This paper presents an efficient OHD‐driven parameter learning scheme for adaptive control, where a variable forgetting factor is specifically designed and is equipped with an estimation error feedback such that exponential parameter convergence is achieved under the SE condition without the aforesaid drawbacks. The proposed parameter learning scheme is incorporated with direct adaptive control to construct an OHD‐based composite learning control strategy. Numerical results have verified the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

Summary: Dynamic regressor extension and mixing (DREM) is a new technique for parameter estimation that has proven instrumental in the solution of several open problems in system identification and adaptive control. A key property of the estimator is that, by generation of scalar regression models, it guarantees monotonicity of each element of the parameter error vector that is a much stronger property than monotonicity of the vector norm, as ensured with classical gradient or least‐squares estimators. On the other hand, the overall performance improvement of the estimator is strongly dependent on the suitable choice of certain operators that enter in the design. In this paper, we investigate the impact of these operators on the convergence properties of the estimator in the context of identification of linear single‐input single‐output time‐invariant systems with periodic excitation. The most important contribution is that the DREM (almost surely) converges under the same persistence of excitation (PE) conditions as the gradient estimator while providing improved transient performance. In particular, we give some guidelines how to select the DREM operators to ensure convergence under the same PE conditions as standard identification schemes. [ABSTRACT FROM AUTHOR]

This paper deals with the problem of identification of photovoltaic arrays' maximum power extraction point-information that is encrypted in the current-voltage characteristic equation. We propose a new parameterisation of the classical 5 parameter models of this function that, combined with the recently introduced identification technique of dynamic regressor extension and mixing, ensures a fast and accurate estimation of all unknown parameters. A concavity property of the current-voltage characteristic equation is then exploited to directly identify the desired voltage operating point. Realistic numerical examples via computer simulations are presented to assess the performance of the proposed approach. [ABSTRACT FROM AUTHOR]

The capacity to learn is one of the fundamental features of autonomous intelligent behavior, which is reflected by parameter convergence in adaptive control.[1] Parameter convergence is desirable as it enhances stability and robustness properties of adaptive control systems such as superior trajectory tracking, accurate online modeling, and robustness against various perturbations,[2] where these properties mainly result from the closed-loop exponential stability. Usually, adaptive control only achieves asymptotic convergence of tracking errors and does not guarantee parameter convergence in the absence of a well-known condition termed persistent excitation (PE). I Concurrent learning i is an emerging technique that makes use of regressor extension and data memory to achieve parameter convergence in adaptive control.[5] In this technique, a dynamic data stack is built to record online historical data (OHD) discretely, and exponential parameter convergence is achieved while sufficiently rich data are recorded to satisfy a condition termed interval excitation (IE), which is strictly weaker than PE. On the other hand, the article by Vedyakov et al[22] applies DREM to achieve continuous-time online frequency estimation for a biased damped sinusoidal signal, where two simple excitation relaxation methods, which make use of integration and transport delay, respectively, are proposed to establish necessary excitation for the first-order gradient-based estimator. [Extracted from the article]

This methodology translates the state estimation problem into a parameter estimation problem and puts it into the indirect adaptation context. Adaptive control had its peak in 1970-1990 when multiple classic results on closed-loop control and parameter estimation were reported, such as References 1-3, and it is still of interest in our time. Their paper "An algebraic, distributed state observer for continuous- and discrete-time linear time-invariant systems with time-varying communication graphs" applies the generalized parameter estimation-based observer methodology. [Extracted from the article]

In this paper, a new class of estimators for permanent magnet synchronous motors is proposed. Using a novel representation of the permanent magnet synchronous motor dynamics and some suitable filtering, we obtain new solutions to two important problems: (a) estimation of the stator resistance and inductance and (b) flux estimation with known electrical parameters. Neither one of the schemes requires knowledge of the mechanical parameters nor of the magnetic flux constant. Conditions for global convergence are given in terms of persistency of excitation requirements that, for the flux observation problem, is imposed only to the rotor speed. Both stability proofs require the availability of rotor speed. However, for the flux observer, extensive simulations using the speed estimate obtained from the estimator (instead of the actual rotor speed) illustrate the robustness of the proposed scheme vis-à-vis this assumption. Simulations also show that the new flux observer outperforms other existing designs. Copyright ©2015 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

Problem of frequency identification performance improvement for a single-tone sinusoidal signal is solved via construction of an adaptive filters cascade. The cascade consists of adaptive band-pass filters tuned by estimates of the frequency provided by a given identification algorithm. Stability of the cascade is studied and boundedness of trajectories is proven with Lyapunov analysis under certain assumption on identification algorithm. Numerical simulations are given illustrating improved identification performance for different identification algorithms. Copyright © 2015 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

In a recent paper it is claimed that adding a filter at the plant input to a classical direct model reference adaptive control constitutes a 'new architecture' for adaptive control. The purpose of this note is to show that this claim is specious. Towards this end, we bring to the readers attention the some recent results where a rigorous mathematical analysis shows that the resulting control signal always converges to the output of a linear time-invariant proportional plus integral (PI) controller, whose gains are independent of the plant parameters. Moreover, it is shown that if the PI controller does not stabilize the plant, the proposed controller will not stabilize it either, making unnecessary the addition of the adaptation. Copyright © 2015 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]