Search

Regularization networks are nonparametric estimators obtained from the application of Tychonov regularization or Bayes estimation to the hypersurface reconstruction problem. Their main drawback back is that the computation of the weights scales as O(n/sup 3/) where n is the number of data. In this paper, we show that for a class of monodimensional problems, the complexity can be reduced to O(n) by a suitable algorithm based on spectral factorization and Kalman filtering. Moreover, the procedure applies also to smoothing splines.

International audience; Recent advances of experimental techniques in biology have led to the production of enormous amounts of data on the dynamics of genetic regulatory networks. We present an approach for the identification of piecewise-affine (PWA) models of genetic regulatory networks from experimental data, focusing on the reconstruction of switching thresholds associated with regulatory interactions. In particular, our method takes into account geometric constraints specific to models of genetic regulatory networks. We show the feasibility of our approach by the reconstruction of switching thresholds in a PWA model of the carbon starvation response in the bacterium Escherichia coli..

In this paper we consider the problem of computing sets of observable states for discrete-time, piecewise affine systems. When the maximal set of observable states is full-dimensional, we provide an algorithm for reconstructing it up to a zero measure set. The core of the method is a quantifier elimination procedure that, in view of basic results on piecewise linear algebra, can be performed via the projection of polytopes on subspaces. We also provide a necessary condition on the minimal length of the observability horizon in order to expect a full-dimensional set of observable states. Numerical experiments highlight that the new procedure is considerably faster than the one proposed in (Ferrari-Trecate and Gati, 2004).

The present paper is devoted to the study of average consensus problems for undirected networks of dynamic agents having communication delays. The accent is put here on the study of the time-delays influence: both constant and time-varying delays are considered, as well as uniform and non uniform repartitions of the delays in the network. The main results provide sufficient conditions (also necessary in most cases) for existence of average consensus under bounded, but otherwise unknown, communication delays. Simulations are provided that show adequation with these results., Comment: 22 pages, 7 figures. Correction of Theorem 7

http://dx.doi.org/10.1007/11730637_16; Recent advances of experimental techniques in biology have led to the production of enormous amounts of data on the dynamics of genetic regulatory networks. In this paper, we present an approach for the identification of PieceWise-Affine (PWA) models of genetic regulatory networks from experimental data, focusing on the reconstruction of switching thresholds associated with regulatory interactions. In particular, our method takes into account geometric constraints specific to models of genetic regulatory networks. We show the feasibility of our approach by the reconstruction of switching thresholds in a PWA model of the carbon starvation response in the bacterium Escherichia coli.

This paper presents an optimal control scheme for a real-time feedback control rate-monotonic scheduling (FC-RMS) system. We consider two-version tasks composed of a mandatory and an optional part to be scheduled according to the FC-RMS. In FC-RMS, the controller provides a feedback strategy for deciding about the execution or rejection of the optional sub-tasks. By modeling the task execution times as random variables, we first find the statistical model of FC-RMS and then we design a pure optimal controller and an optimal controller with feedforward integral compensation. The comparison of these two schemes with common Proportional-Integral-Derivative (PID) controller highlights the benefit of the optimal scheme with integral compensation. The results are demonstrated through the real implementation of FC-RMS on RT-Linux.

In This work, three recently proposed procedures for the identification of piece-wise autoregressive exogenous (PWARX) models are compared. Quantitative measures for the quality of the obtained models are proposed. Using one dimensional examples, specific behaviors of the methods are pointed out. An experimental example is considered as well.

When performing regression with piecewise affine maps, the most challenging task is to classify the data points, i.e. to correctly attribute a data point to the affine submodel that most likely generated it. In this paper, we consider a regression scheme similar to the one proposed in~\cite{FMLMa01,FMLM03} that reduces the classification step to a clustering problem in presence of outliers. However, instead of the K-means procedure adopted in~\cite{FMLMa01,FMLM03}, we propose the use of single-linkage clustering that estimates automatically the number of submodels composing the piecewise affine map. %on the basis of a threshold whose proper value depends on the %distance between model coefficients and outliers. Moreover we prove that, under mild assumptions on the data set, single-linkage clustering can guarantee optimal classification in presence of bounded noise.

A new connectionist model for the solution of piecewise lin- ear regression problems is introduced; it is able to reconstruct both con- tinuous and non continuous real valued mappings starting from a finite set of possibly noisy samples. The approximating function can assume a different linear behavior in each region of an unknown polyhedral parti- tion of the input domain. The proposed learning technique combines local estimation, clustering in weight space, multicategory classification and linear regression in order to achieve the desired result. Through this approach piecewise affine solutions for general nonlinear regression problems can also be found.

A new learning algorithm for solving piecewise linear regression problems is proposed. It is able to train a proper multilayer feedforward neural network so as to reconstruct a target function assuming a different linear behavior on each set of a polyhedral partition of the input domain. The proposed method combine local estimation, clustering in weight space, classification and regression in order to achieve the desired result. A simulation on a benchmark problem shows the good properties of this new learning algorithm.

In this paper we focus on the identification of discrete- time hybrid systems in the Piece-Wise Affine (PWA) form. This problem can be formulated as the recon- struction of a possibly discontinuous PWA map with a multi-dimensional domain. In order to achieve our goal, we propose an algorithm that exploits the com- bined use of clustering; linear identification, and clas- sification techniques. This allows to identlfy both the affine submodels and the polyhedral partition of the do- main on which each submodel is valid.

We propose a new technique for the identification of discrete-time hybrid systems in the piecewise affine (PWA) form. This problem can be formulated as the reconstruction of a possibly discontinuous PWA map with a multi-dimensional domain. In order to achieve our goal, we provide an algorithm that exploits the combined use of clustering, linear identification, and pattern recognition techniques. This allows to identify both the affine submodels and the polyhedral partition of the domain on which each submodel is valid avoiding gridding procedures. Moreover, the clustering step (used for classifying the datapoints) is performed in a suitably designed feature space which allows also to reconstruct different submodels that share the same coefficients but are designed on different regions. Measures of confidence on the samples are introduced and exploited in order to improve the performance of both the clustering and the final linear regression procedure.

Mixed Logical Dynamical (MLD) systems are introduced as a new system type. The MLD form is capable to model a broad class of systems arising in many applications: linear hybrid systems; sequential logical systems (finite; state machines, automata): nonlinear dynamic systems, where the nonlinearity can be expressed through combinational logic: some classes of discrete event systems; constrained linear systems. Controllability/verification and observability of MLD systems arid other system theoretic properties are defined. Tests for these properties are formulated in the form of Mixed-Integer Linear Programs. Moving horizon control and estimation strategics with stability guarantees are proposed. These strategies require the iterative solution of Mixed-Integer Quadratic Programs. Several examples communicate the power and versatility of the proposed framework.

Experimental techniques in molecular biology have led to the production of enormous amounts of data on the dynamics of cellular processes. The availability of time series data characterizing genomic, proteomic and metabolic systems must be complemented with formal methods for identifying quantitative models of networks of interactions. Reverse-engineering of regulatory networks is a central issue in modern biology because, beside enabling the computer-based simulation of biological systems, it promotes the understanding of cell functioning and underlies the design of interventions of biotechnological or biomedical relevance. However, standard system identification techniques are unlikely to work out of the box since they must cope with (1) the complexity and the high nonlinearity of biological systems; (2) the quality and type of available biological data; (3) the stochastic nature of chemical interactions; and (4) the interaction of discrete events and continuous dynamics. The aim of this special issue is to present some very recent achievements in system identification tailored to the reconstruction of biological processes.

In this note, we introduce the framework of partial difference equations (PdEs) over graphs for analyzing the behavior of multi-agent systems equipped with decentralized control schemes. Both leaderless and leader-follower models are considered. PdEs mimic partial differential equations (PDEs) on graphs and can be studied by introducing concepts of functional analysis strongly inspired to the corresponding ones arising in PDEs theory. We generalize different models proposed in the literature by introducing errors in the agent dynamics and analyze agent coordination through the joint use of PdEs and automatic control tools. Moreover, for the simplest control schemes, we show that the resulting PdEs enjoy properties that are similar to those of well-known PDEs like the heat equation, thus allowing to exploit physical-based reasoning for conjecturing formation properties.

This paper defines a Maximum Likelihood Estimator (MLE) for the admittance matrix estimation of distribution grids, utilising voltage magnitude and power measurements collected only from common, unsychronised measuring devices (Smart Meters). First, we present a model of the grid, as well as the existing MLE based on voltage and current phasor measurements. Then, this problem formulation is adjusted for phase-less measurements using common assumptions. The effect of these assumptions is compared to the initial problem in various scenarios. Finally, numerical experiments on a popular IEEE benchmark network indicate promising results. Missing data can greatly disrupt estimation methods. Not measuring the voltage phase only adds 30\% of error to the admittance matrix estimate in realistic conditions. Moreover, the sensitivity to measurement noise is similar with and without the phase., Comment: 5 pages, Grid Edge 2023 conference

The paper considers the problem of detecting cyber-attacks occurring in communication networks typically used in the secondary control layer of DC microgrids. The proposed distributed methodology allows for scalable monitoring of a microgrid and is able to detect the presence of data injection attacks in the communications among Distributed Generation Units (DGUs) - governed by consensus-based control - and isolate the communication link over which the attack is injected. Each local attack detector requires limited knowledge regarding the dynamics of its neighbors. Detectability properties of the method are analyzed, as well as a class of undetectable attacks. Some results from numerical simulation are presented to demonstrate the effectiveness of the proposed approach.

This paper deals with a novel Plug-and-Play (PnP) architecture for the control and monitoring of Large-Scale Systems (LSSs). The proposed approach integrates a distributed Model Predictive Control (MPC) strategy with a distributed Fault Detection (FD) architecture and methodology in a PnP framework. The basic concept is to use the FD scheme as an autonomous decision support system: once a fault is detected, the faulty subsystem can be unplugged to avoid the propagation of the fault in the interconnected system. Analogously, once the issue has been eventually solved, the disconnected subsystem can be re-plugged-in. PnP design of local controllers and detectors allow these operation to be performed safely, i.e. without spoiling stability and constraint satisfaction for the whole LSS. The PnP distributed MPC is derived for a class of nonlinear LSS and an integrated PnP distributed FD architecture is proposed. Early simulation results show the effectiveness and potentialities of the general methodology.

In this paper we address the problem of driving a group of agents towards a consensus point when the agents have a discrete-time single- or double-integrator dynamics and the communication network is time-varying. We propose decentralized Model Predictive Control (MPC) schemes that take into account constraints on the agents' input and show that they guarantee consensus under mild assumptions. Since the global cost does not decrease monotonically, it cannot be used as a Lyapunov function for proving convergence to consensus. For this reason, our proofs exploit geometric properties of the optimal path followed by individual agents.

In this paper the load optimization of a combined cycle power plant under consideration of the real cost of lifetime usage is accomplished by exploiting hybrid systems, i.e., systems evolving according to continuous dynamics, discrete dynamics, and logic rules. The possibility of turning on/off the gas and steam turbines, the operating constraints (minimum up and down times) and the different types of start up of the turbines characterize the hybrid behavior of a combined cycle power plant. In order to model both the continuous/discrete dynamics and the switching between different operating conditions we use the framework of Mixed Logic Dynamical systems. Next, we recast the economic optimization problem as a Model Predictive Control (MPC) problem, that allows us to optimize the plant operations by taking into account the time variability of both prices and electricity/steam demands. Because of the presence of integer variables, the MPC scheme is formulated as a mixed integer linear program that can be solved in an efficient way by using commercial solvers.

The paper considers the problem of detecting cyber-attacks occurring in communication networks for distributed control schemes. A distributed methodology is proposed to detect the presence of malicious attacks aimed at compromising the stability of large-scale interconnected systems and multi-agent systems governed by consensus-based controllers. Only knowledge of the local model is required. The detectability properties of the proposed method are analyzed. A class of undetectable attacks is identified. Preliminary simulation results show the effectiveness of the proposed approach.

The Wold decomposition of discrete-time CS (cyclostationary) processes into a regular and predictable component is considered. Using the properties of the zeros of linear periodic systems, it is shown that the predictable component consists of a linear combination of sine waves.

The problem of replay attacks in the communication network between Distributed Generation Units (DGUs) of a DC microgrid is examined. The DGUs are regulated through a hierarchical control architecture, and are networked to achieve secondary control objectives. Following analysis of the detectability of replay attacks by a distributed monitoring scheme previously proposed, the need for a watermarking signal is identified. Hence, conditions are given on the watermark in order to guarantee detection of replay attacks, and such a signal is designed. Simulations are then presented to demonstrate the effectiveness of the technique.

This paper deals with a novel Plug-and-Play (PnP) architecture for the control and monitoring of Large-Scale Systems (LSSs). The proposed approach integrates a distributed Model Predictive Control (MPC) strategy with a distributed Fault Detection (FD) architecture and methodology in a PnP framework. The basic concept is to use the FD scheme as an autonomous decision support system: once a fault is detected, the faulty subsystem can be unplugged to avoid the propagation of the fault in the interconnected LSS. Analogously, once the issue has been solved, the disconnected subsystem can be re-plugged-in. PnP design of local controllers and detectors allow these operations to be performed safely, i.e. without spoiling stability and constraint satisfaction for the whole LSS. The PnP distributed MPC is derived for a class of nonlinear LSS and an integrated PnP distributed FD architecture is proposed. Simulation results show the effectiveness and the potential of the general methodology.

In this paper, the short-term scheduling optimization of a combined cycle power plant is accomplished by exploiting hybrid systems, i.e., systems evolving according to continuous dynamics, discrete dynamics, and logic rules. Discrete features of a power plant are, for instance, the possibility of turning on/off the turbines, operating constraints like minimum up and down times and the different types of start up of the turbines. On the other hand, features with continuous dynamics are power and steam output, the corresponding fuel consumption, etc. The union of these properties characterize the hybrid behavior of a combined cycle power plant. In order to model both the continuous/discrete dynamics and the switching between different operating conditions, we use the framework of mixed logic dynamical (MLD) systems. Then, we recast the economic optimization problem as a model predictive control (MPC) problem, that allows us to optimize the plant operations by taking into account the time variability of both prices and electricity/steam demands. Because of the presence of integer variables, the MPC scheme is formulated as a mixed integer linear program that can be solved in an efficient way via dedicated software.

The problem of driving a collection of mobile robots to a given target location is studied in the context of partial difference equations. In particular, we are interested in achieving this transfer while ensuring that the agents stay in the convex polytope spanned by dedicated leader-agents, whose dynamics will be given by a hybrid Stop-Go policy. The resulting system ensures containment through the enabling result that under a Laplacian, decentralized control strategy for the followers, these followers will converge to a location in the convex leader polytope, as long as the leaders are stationary and the interaction graph is connected. Simulation results testify to the viability of the proposed, hybrid control strategy.

In this paper we propose a new decentralized control scheme for Islanded microGrids (ImGs) composed by the interconnection of Distributed Generation Units (DGUs). Local controllers regulate voltage and frequency at the Point of Common Coupling (PCC) of each DGU and they are able to guarantee stability of the overall ImG. The control design proce- dure is decentralized, since, besides two global scalar quantities, the synthesis of a local controller uses only information on the corresponding DGU and lines connected to it. Most important, our design procedure enables Plug-and-Play (PnP) operations: when a DGU is plugged in or out, only DGUs physically connected to it have to retune their local controllers. We study the performance of the proposed controllers simulating different scenarios in MatLab/Simulink and using indexes proposed in IEEE standards.

In this work we propose several algorithms to solve the reconfiguration problem for linear and hybrid systems. In particular, we consider the decision about the usage of redundant hardware in order to compensate for faults. While this problem can be translated into a constrained model predictive control framework, the computational complexity grows very fast, as the number of possible decisions increases. In this work we propose schemes, that require low computational effort. We discuss the applicability of the methods considering the reconfiguration of the three tank benchmark system.

In this second part of the paper we exploit the framework of Partial difference Equations (PdEs) over graphs for analyzing the behavior of multi-agent systems equipped with potential field based control schemes. We consider agent dynamics affected by errors and model the collective dynamics through nonlinear PdEs. Hinging on the properties of the Laplacian operator on graph, discussed in Part I, we prove alignment and collision avoidance both in leaderless and leader-follower models.

The increasing penetration of intermittent distributed energy resources in power networks calls for novel planning and control methodologies which hinge on detailed knowledge of the grid. However, reliable information concerning the system topology and parameters may be missing or outdated for temporally varying electric distribution networks. This brief proposes an online learning procedure to estimate the network admittance matrix capturing topological information and line parameters. We start off by providing a recursive identification algorithm exploiting phasor measurements of voltages and currents. With the goal of accelerating convergence, we subsequently complement our base algorithm with a design-of-experiment procedure which maximizes the information content of data at each step by computing optimal voltage excitations. Our approach improves on existing techniques, and its effectiveness is substantiated by numerical studies on a modified IEEE testbed.

We consider the problem of dispatching Wind-Farm (WF) power demand to individual Wind Turbines (WTs) with the goal of minimizing mechanical stresses. We assume wind is strong enough to let each WT produce the required power and propose different closed-loop Model Predictive Control (MPC) dispatching algorithms. Similarly to existing approaches based on MPC, our methods do not require changes in WT hardware but only software changes in the SCADA system of the WF. However, differently from other MPC schemes, we augment the model of a WT with an ARMA predictor of the wind turbulence, which reduces uncertainty in wind predictions over the MPC control horizon. This allows us to develop both stochastic and deterministic MPC algorithms. In order to compare different MPC schemes and demonstrate improvements over classic open-loop schedulers, we use simulations based on the SimWindFarm toolbox for MatLab.

When tuning the smoothness parameter of nonparametric regression splines, the evaluation of the so-called degrees of freedom is one of the most computer-intensive tasks. In the paper, a closed-form expression of the degrees of freedom is obtained for the case of cubic splines and equally spaced data when the number of data tends to infinity. State-space methods, Kalman filtering and spectral factorization techniques are used to prove that the asymptotic degrees of freedom are equal to the variance of a suitably defined stationary process. The closed-form expression opens the way to fast spline smoothing algorithms whose computational complexity is about one-half of standard methods (or even one-fourth under further approximations).

We consider the application of passivity theory to the problem of voltage stabilization in DC microgrids (DCmGs) given by the interconnection of Distributed Generation Units (DGUs), dynamic RL lines, and loads. We first provide a novel result on stable interconnection of multiple passive systems and then apply it to DCmGs. More specifically, we consider the decentralized controllers proposed in [1] synthesized by solving linear matrix inequalities, which allow addition and removal of DGUs in a plug-n-play fashion. We show that they passivate DGUs and combine this property with the inherent passivity of RL lines for proving voltage stability without Quasi-Stationary-Line approximations, which are assumed in [1]. Theoretical results are backed up by simulations in PSCAD.

In this paper we consider discrete-time piecewise affine hybrid systems with boolean inputs, outputs and states and show that they can be represented in a logic canonical form where the logic variables influence the switching between different submodels but not the continuous-valued dynamics. We exploit this representation for studying Lagrange stability and developing performance analysis procedures based on linear matrix inequalities. Moreover, by using arguments from dissipativity theory for nonlinear systems, we generalize our approach to solve the H-infinity analysis problem.

In this paper we present various algorithms both for stability analysis and state-feedback design for discrete-time piecewise affine systems. As in (Johansson and Rantzer, 1998), our approach hinges on the use of piecewise quadratic Lyapunov functions that can be computed as the solution of a set of linear matrix inequalities. We show that the continuity of the Lyapunov function is not required in the discrete-time case. Moreover, the basic algorithms are made less conservative by exploiting the switching structure of piecewise affine systems and by using relaxation procedures.

In this paper, we provide an extension of the scalable algorithm proposed in [1] for the design of Plug-and-Play (PnP) controllers for AC Islanded microGrids (ImGs). The method in [1] assumes Distributed Generation Units (DGUs) are arranged in a load-connected topology, i.e. loads can appear only at the output terminals of inverters. For handling totally general interconnections of DGUs and loads, we describe an approach based on Kron Reduction (KR), a network reduction method giving an equivalent load-connected model of the original ImG. However, existing KR approaches can fail in preserving the structure of transfer functions representing transmission lines. To avoid this drawback, we introduce an approximate KR algorithm, still capable to represent exactly the asymptotic periodic behavior of electric signals in balanced RL line networks with balanced and unbalanced loads. Our results are backed up with simulations illustrating features of the new KR approach as well as its use for designing PnP controllers in a 21-bus ImG derived from an IEEE test feeder.

Bayesian regression, a nonparametric identification technique with several appealing features, can be applied to the identification of NARX (nonlinear ARX) models. However, its computational complexity scales as O(N/sup 3/) where N is the data set size. In order to reduce complexity, the challenge is to obtain fixed-order parametric models capable of approximating accurately the nonparametric Bayes estimate avoiding its explicit computation. In this work we derive, optimal finite-dimensional approximations of complexity O(N/sup 2/) focusing on their use in the parametric identification of NARX models.

In this first part of a two-parts paper we introduce the framework of Partial difference Equations (PdEs) over graphs for analyzing the behavior of multi-agent systems equipped with decentralized control schemes. We generalize the Vicsek's model by introducing errors in the agent dynamics and analyze agent alignment in leaderless and leader-follower models through the joint use of PdEs and automatic control tools. Moreover, we show that the resulting PdEs enjoy properties that are similar to those of well-known Partial Differential Equations (PDEs) like the heat equation, thus allowing to exploit physical-based reasoning for conjecturing properties of the collective dynamics.

In this paper we focus on observability for hybrid systems in the mixed-logic dynamical form. We show that the maximal set of observable states, that is usually nonconvex and disconnected, can be represented as the union of finitely many polytopic regions. The argument, which is based on multi-parametric programming theory, is constructive and provides an algorithm for the computation of the regions.

Bioinformatics, 26 (9), ISSN:1367-4803, ISSN:1460-2059

In this paper, we address the problem of synthesizing decentralized current controllers for grid-feeding converters of current-controlled distributed generation units (CDGUs) in dc microgrids (MGs). Notably, a plug-and-play (PnP) design procedure is proposed to achieve grid-feeding current tracking while preserving the collective MG stability. Through the presented control scheme, seamless addition/removal of each CDGU to/from the MG is ensured, with no need to update controllers of neighboring CDGUs and to know the information of the MG. At the mathematical level, the set of control coefficients guaranteeing the aforementioned features is explicitly characterized in terms of simple inequalities. The inequality set only depends on the local parameters. Moreover, the proof of the MG closed-loop stability exploits structured Lyapunov functions, the LaSalle invariance theorem and properties of graph Laplacians. Finally, theoretical results are validated by hardware-in-loop simulation tests.

Kron reduction (KR) is a methodology for analyzing an electrical network by replacing it with a simpler circuit having less nodes but the same terminal behavior of voltages and currents at target vertices. Existing approaches to instantaneous KR, however, can fail in preserving the structure of transfer functions representing power lines. Therefore, even if the original lines are passive $RLC$ circuits, reduced lines might have transfer functions that do not correspond to a physically realizable passive system. To overcome this drawback, in this paper we focus on $RL$ line models and propose two approximate KR algorithms producing reduced lines with the first-order transfer functions and capable of representing exactly the asymptotic behavior of electric signals, even if they are unbalanced. Then, we show how to apply these KR methods to the design of decentralized voltage and frequency controllers for AC islanded microgrids. Notably, we focus on the plug-and-play algorithm previously proposed by some of the authors, which assumes loads connected to the inverter outputs, and generalize it to networks where loads appear in arbitrary positions. Theoretical results are validated with numerical examples and the application of KR for designing PnP controllers is assessed through a simulation on a 21-bus microgrid.

This paper presents methods for the parameter identification of a model of subtilin production by Bacillus subtilis. Based on a stochastic hybrid model, identification is split in two subproblems: estimation of the genetic network regulating subtilin production from gene expression data, and estimation of population dynamics based on nutrient and population level data. Techniques for identification of switching dynamics from sparse and irregularly sampled observations are developed and applied to simulated data. Numerical results are provided to show the effectiveness of our methods.

This paper performs an extensive review on control schemes and architectures applied to dc microgrids (MGs). It covers multilayer hierarchical control schemes, coordinated control strategies, plug-and-play operations, stability and active damping aspects, as well as nonlinear control algorithms. Islanding detection, protection, and MG clusters control are also briefly summarized. All the mentioned issues are discussed with the goal of providing control design guidelines for dc MGs. The future research challenges, from the authors’ point of view, are also provided in the final concluding part.

In this paper we propose a novel partition-based state estimator for linear discrete-time systems composed of physically coupled subsystems affected by bounded disturbances. The proposed scheme is distributed in the sense that each local state estimator exploits suitable pieces of information from parent subsystems. Moreover, differently from schemes based on moving horizon estimation, it does not require the on-line solution to optimization problems. Our method also guarantees the satisfaction of constraints on local estimation errors. We achieve our aims exploiting the notion of practical robust positive invariance developed in (Rakovic et al, 2012). As an example, we illustrate the use of the distributed state estimator for reconstructing the states of a power network system.

This paper concerns parameter identification for the class of jump Markov linear systems. We present recursive formulae for the computation of the second-order statistics of a jump Markov process and discuss their application to the estimation of the model parameters given data from a single or repeated experiments. Simulation results are provided to show the effectiveness of our approach.

We propose a state estimator for linear discrete-time systems composed by coupled subsystems affected by bounded disturbances. The architecture is distributed in the sense that each subsystem is equipped with a local state estimator that exploits suitable pieces of information from parent subsystems. Furthermore, each local estimator reconstructs the state of the corresponding subsystem only. Differently from methods based on moving horizon estimation, our approach does not require the on-line solution to optimization problems. Our state-estimation scheme, which is based on the notion of practical robust positive invariance developed in (Rakovic et al., 2011), also guarantees satisfaction of constraints on local estimation errors and it can be updated with a limited computational effort when subsystems are added or removed.

We propose a new decentralized control scheme for DC Islanded microGrids (ImGs) composed by several Distributed Generation Units (DGUs) with a general interconnection topology. Each local controller regulates to a reference value the voltage of the Point of Common Coupling (PCC) of the corresponding DGU. Notably, off-line control design is conducted in a Plug-and-Play (PnP) fashion meaning that (i) the possibility of adding/removing a DGU without spoiling stability of the overall ImG is checked through an optimization problem; (ii) when a DGU is plugged in or out at most neighbouring DGUs have to update their controllers and (iii) the synthesis of a local controller uses only information on the corresponding DGU and lines connected to it. This guarantee total scalability of control synthesis as the ImG size grows or DGU gets replaced. Yes, under mild approximations of line dynamics, we formally guarantee stability of the overall closed-loop ImG. The performance of the proposed controllers is analyzed simulating different scenarios in PSCAD., Comment: arXiv admin note: text overlap with arXiv:1405.2421