Search

In this paper, a resilient distributed control scheme against covert attacks for constrained multi-agent networked systems is developed. The idea consists in an adequate deployment of predictive arguments with a twofold aim: detection of malicious agent behaviors and control actions implementation to mitigate as much as possible undesirable knock-on effects.

Predictive control presents an intrinsic capability to mitigate faulty events that can be achieved by properly defining operational constraints and/or running costs. Within this context this work presents a model predictive control (MPC) approach to deal with faults related to loss of actuators effectiveness. These undesired events occur when the associated degradation effects, whose evolution is assumed to be measurable and predictable, overcomes certain thresholds. The presented scheme consists of two modules: a prognostic device in charge of monitoring actuators health and a control reconfiguration unit whose action is based on actuators degradation information. The idea exploits a switching systems paradigm to formally model healthy and faulty plant configurations and to offline determine sequences of pre-computed inner approximations of one-step controllable sets. Such regions are then on-line selected by a switching logic to compute a proper control signal on the basis of the measured degradation level.

In this paper, we present a novel resilient control architecture capable to manage replay attacks for multi-agent discrete-time linear systems subject to input and state constraints. By considering a leader-follower configuration, the basic idea consists of exploiting model predictive control arguments to apply adequate control action in order to isolate the attacked unit that otherwise could compromise system operations. To this end, a set-theoretic receding horizon control strategy is developed that is also capable to instantaneously detect the attacked agent along the platoon chain. Finally, we describe a set of simulations on a group of mobile robots to demonstrate the effectiveness of the proposed approach.

In this paper, we propose a distributed model predictive control strategy for vehicle teams described by linear time-invariant models and organized as leader-follower configurations. Based on mild assumptions on vehicle communication facilities, the obstacle avoidance path planning problem is addressed by developing a parallel scheme which takes advantage of feasibility arguments in order to safely reorganize, whenever necessary, the vehicle team.

In this paper a novel receding horizon control scheme is presented for solving networked leader-follower formation control problems in the presence of time-varying time-delay on the communication channel. The algorithm is based on a reachability analysis where a sequence of nested one-step ahead controllable sets has been derived by taking into account the prescribed state and input constraints, the interconnected nature of the system and the time-delay induced by the network latency. The novelty of the proposed scheme relies on its capability to jointly use sequences of one-step ahead controllable sets and polyhedral state-space partitions in order to compute on-line an admissible and optimal control action via a distributed receding-horizon computational paradigm.

This paper presents a Fault Tolerant Control scheme for constrained discrete time linear systems against stuck actuators and bounded disturbances. The results here proposed are a significant generalization of a similar algorithm which has been designed only to manage healthy/out-of-service in the actuator units. A first key feature is the capability to exploit existing methods based on polyhedral algebra and computational geometry in order to characterize the reconfiguration logic of the proposed fault tolerant architecture. This is achieved by defining a set of a finite number of piecewise affine systems capable to cover all of the admissible faulty stuck scenario. It is then proved that one-step sequences of controllable sets can be formally defined within a piecewise affine system paradigm under the key property that each region is a polygon.

We present a new actuator fault tolerant control strategy for linear, discrete-time systems. The scheme consists of three modules: a bank of estimators, each one associated with healthy and faulty model configurations, a logic mechanism for identifying healthy-to-faulty and faulty-to-healthy transitions, an estimate based control reconfiguration unit. The idea is to abstractly describe healthy and faulty plant configurations by means of a switching system and to make use sequences of pre-computed inner approximations of the one-step controllable sets associated to the obtained model configurations. Such sets are then on-line exploited together with a switching logic to determine the current plant configuration on the basis of the estimate state provided by the observers.

A Model Predictive Control (MPC) strategy for nonlinear Networked Control Systems (NCS) described by a polynomial plants subject to input, state constraints and data-losses is introduced. The results here proposed represent a significant extension to the polynomial framework of a recent Receding Horizon Control (RHC) scheme developed for input-saturated multi-model systems. Sum-of-Squares (SOS) conditions for polynomial systems are derived together with an SDP formulation of the on-line optimization problems.

In this paper reachability/controllability properties for networked systems described by uncertain polytopic linear plants subject to time-varying state delays and input constraints are analyzed. Up to our best acknowledge, previous technical contributions on the matter were not been capable to properly address time-delay occurrences. To this end a new result making it possible to efficiently compute the set of states that can be robustly steered in a finite number of steps, via state feedback control, to a given target set is here proposed. A final simple example is used to show the applicability of the proposed results.

We propose a novel receding horizon strategy for Networked Control Systems described by uncertain polytopic linear plants subject to time-varying delays and data-losses. We make use of sequences of pre-computed inner approximations of the one-step ahead state prediction sets on-line exploited as target sets for state predictions. The present approach is capable to deal with data-loss events of arbitrarily length without compromising closed-loop stability and constraints fulfilment.

The problem of estimating the domain of attraction (DoA) of nonlinear time-delay polynomial systems subject to actuator saturation is considered in this paper via a Sum-of-Squares (SOS) approach. It will be shown that the joint use of sum-of-squares decomposition techniques and semidefinite programming provides an efficient methodology for the analysis and synthesis of polynomial nonlinear systems via Lyapunov stability arguments. Specifically, we present a two-step algorithmic procedure: first, a Lyapunov-Krasovskii functional/controller pair complying with the prescribed constraints is derived and then an estimate of the true DoA is determined. The effectiveness of the proposed framework is illustrated via a numerical example.

In this paper we propose a Fault Detection and Isolation (FDI) filter design method for Spark Injection Engines. Starting from a detailed nonlinear mean-value representation of the engine, a LPV approximation is obtained based on a judicious convex interpolation of a family of linearized models. A LPV-FDI filter based on a bank of Luenberger observers is synthesized by ensuring guaranteed levels of disturbance rejection and fault detection and isolation. The resulting diagnostic setup is parameter-dependent and uses a set of engine parameters, assumed measurable on-line, as a scheduling vector. The effectiveness of the LPV-FDI framework is illustrated by numerical examples where the diagnostic capabilities of the proposed FDI architecture are proved.

In this paper we propose a Fault Detection and Isolation (FDI) filter design method for Spark Injection Engines. Starting from a detailed nonlinear Mean-value engine mathematical representation, a LPV approximation based on a judicious convex interpolation of a family of linearized models is obtained. An LPV-FDI filter based on the Luenberger observer theory is synthesized by ensuring guaranteed levels of disturbance rejection and fault detection and isolation. The resulting diagnostic filter is parameter-dependent and uses a set of scheduling engine parameters, assumed measurable on-line. The effectiveness of the LPV-FDI framework is illustrated by numerical examples. The obtained LPV approximation is here validated and the diagnostic capabilities of the proposed FDI architecture proved.

A novel nonlinear Luenberger-like filter design method is discussed for fault detection of nonlinear polynomial systems subject to unknown inputs. The disturbance rejection/ fault sensitivity conditions are characterized by exploiting Algebraic Geometry methods. By means of the Positivstellensatz Theorem, a Sum of Squares (SOS) formulation of the design problem is obtained and an adaptive threshold is proposed to reduce the effect of false alarms. The effectiveness of the filter is illustrated via a numerical example taken from the literature.

This paper presents a novel Model Predictive Control strategy for input-saturated nonlinear systems having a polynomial structure. The method is based on algebraic geometry and viability theory arguments for computing the sets of states that can be steered in a finite number of steps, via a general control law, to a given robust positive invariant set.

An off-line Model Predictive Control (MPC) method based on ellipsoidal calculus and viability theory is described in order to address feedback regulation problems for input-saturated multi-model Linear Parameter Varying (LPV) systems. In order to reduce the computational burdens and the conservativeness of traditional robust MPC schemes, the proposed approach makes use of a scheduling linear control law which is based on a tighter refinement of the state trajectories tube. The control inputs are on-line obtained as the solutions of simple and numerically low-demanding optimization problems subject to set-membership constraints. The measurable plant parameters are directly used for feedback purposes. A numerical example is presented in order to show the benefits of the proposed regulation strategy and contrasted with a robust paradigm (the LPV hypothesis is not taken into account) and with a fast MPC algorithm for LPV systems which exploits a more traditional and conservative refinement of the state trajectory tube.

In this paper we propose a Fault Detection (FD) filter design method for linear parameter varying (LPV) systems. The FD filter is an optimal Hinfin Luenberger observer synthesized by minimizing frequency conditions which ensure guaranteed levels of disturbance rejection and fault detection. Via the Bounded Real Lemma (BRL) and the Separation Principle the design method is formulated as a convex LMI optimization problem. The resulting residual generator is parameter-dependent and uses the plant parameter assumed measurable on-line. Finally, a FD threshold logic is proposed in order to reduce the generation of false alarms. The effectiveness of the design technique is illustrated via a numerical example.

A novel model predictive control (MPC) strategy for input-saturated polytopic linear parameter varying (LPV) discrete-time systems is proposed. It is assumed that the scheduling vector of the plant parametrization is available at each time instant. A new dilation technique and parameter varying Lyapunov functions are exploited to reduce conservativeness of existing robust MPC paradigms. The contribution of this paper is to extend the RHC scheme presented in [5] to the general case of control horizons of arbitrary length N. Feasibility and closed-loop stability of this strategy are proved and two final numerical examples presented in order to show performance improvements w.r.t. preexisting techniques.

New stabilizability conditions for input-saturated multi-model Linear Parameter Varying (LPV) discrete-time systems are proposed. They are obtained by means of a "refinement" of the usual quadratic stabilizability conditions. It is proven that such conditions can be made significantly more tight by a better characterization of the politopic set which embeds the true LPV parameter realization. An application to a computationally low demanding "off-line" Receding Horizon Control (RHC) scheme is provided. A numerical example is also presented in order to show that the regulation strategy based on the refined conditions significantly improves the performance of the control strategy and its basin of attraction.

A novel predictive control strategy for input- saturated Norm-Bounded LPV discrete-time systems is proposed. The solution is computed by minimizing an upper-bound to the "worst-case" infinite horizon quadratic cost under the constraint of steering the future state evolutions, emanating from the current state, into a feasible and positive invariant set. It will be shown that the "size" of this terminal set depends on the rate of changes of the scheduling parameter which is assumed bounded and measurable.

This paper presents a new dilated LMI approach to constrained MPC for LPV discrete-time systems. The main idea exploited in the proposed algorithm consists of using both the dilation argument and the direct measure of the LMI parameter at each time step. The effectiveness of the technique is compared to preexisting techniques by means of a numerical example.

This paper presents a procedure for allocating measurement sensors on a rectangular plate, to achieve deformation measurement. The aim is to obtain a good reconstruction of the plate's state that enable us to reduce its vibrations through a state-feedback control law.

The results of some numerical experiments performed on a particular reaction-diffusion system are presented and discussed. They show that, at particular values of some parameters, a transition occurs from a periodic to an aperiodic behavior in time, to which there corresponds the onset of a very complex spatial patterns. >

A novel robust predictive control algorithm is presented for input-saturated uncertain discrete-time linear systems described by structured norm-bounded uncertainties. The solution is based on the minimization, at each time instant, of a LMI convex optimiza- tion problem obtained by a recursive use of the S-procedure. Spe- cific features of this scheme are that is based on "closed-loop" predictions and that the ellipsoidal terminal constraint can, op- tionally. be updated on-line. The general case of N free control moves is presented. Stability and feasibility are proved and com- parisons with robust multi-model (polytopic) MPC algorithms are reported.

A robust MPC strategy is described for solving set-point problems for a class of nonlinear systems subject pointwise-in-time input and/or state related constraints. The trajectories of the nonlinear system are embedded within those of a linear parameter varying (LPV) uncertain system so that the convex computational paradigm of the LMI optimization can be used. In this note, the main features of a convex robust linear MPC scheme, suitable for the embedding, are presented and it is contrasted, via a highly nonlinear example, with a direct nonlinear non-convex MPC scheme in terms of control performance and computational burdens.

This paper presents a novel solution to the fault detection and isolation observer design problem for linear time-invariant systems. A gradient flow approach is proposed for synthesizing a residual generator under optimal eigenstructure assignment. This is achieved by minimizing the spectral condition number of the observer eigenvector matrix. The properties of convergence of the gradient flow solution are proved and its efficiency demonstrated via a numerical example. Copyright © 2007 John Wiley & Sons, Ltd.

A novel robust predictive control algorithm is presented for input-saturated uncertain discrete-time linear systems described by structured norm-bounded uncertainties. The solution is based on the minimization, at each time instant, of an LMI convex optimization problem obtained by a recursive use of the S-procedure. Stability and feasibility are proved and comparisons with robust multi-model (polytopic) MPC algorithms are reported.

For linear systems with polytopic uncertainties, the problem of robust finite-time stabilization is reduced to a system of linear matrix inequalities.

Most robot tactile sensing research has focused on object shape recognition with tactile sensors. In this paper we examine the model of a tactile pad sensor, assumed to be a finite dimension medium. The model of the tactile pad sensor is obtained using an integral relation between the boundary conditions and the stress tensor. A set of stress sensors is allocated in the tactile pad, the model is discretized and a direct relation between the parameters of the shape to be recognized and the measured stress is obtained. Due to the ill-posedness of the problem, classical least squares estimation and regularized estimation are examined. Numerical simulations in order to evaluate the inversion technique with different shapes of the object, are presented.

Model predictive control strategies are described for solving set-point problems for nonlinear systems when pointwise-in-time input and/or state related inequality constraints have to be fulfilled. The trajectories of the nonlinear system are embedded within those of an auxiliary norm-bounded uncertain linear system in order to use simpler computational paradigms based on convex programming (LMIs). These MPC schemes are based on the minimization, at each time instant, of an upper bound on the worst case quadratic performance index. The conditions on the upper bound are obtained by means of the S-procedure. A control strategy based on N control free moves plus a feedback gain matrix is used. In order to demonstrate the effectiveness of the solution and the computational complexity reduction achievable, a numerical experiment is presented and contrasted to standard robust multi-model (polytopic) MPC schemes.

In this article, we introduce a robust nonfragile state feedback controller which is also optimal with respect to a quadratic performance index, using linear matrix inequalities (LMI). The uncertainties are assumed to be polytopic, both in the controller gains and the system dynamics. A numerical example is presented to demonstrate the efficiency of this method, and the controller turns out to be robust with respect to the uncertainties in the plant and the controller.

This paper describes the synthesis of non-fragile or resilient regulators for linear systems. A general framework for fragility is described using state-space methodologies, and the LQ/H2 static state-feedback problem is examined in detail. We discuss the multiplicative structured uncertainties case, and propose remedies of the fragility problem using an optimization programming framework via matrix inequalities. A special case that leads to a convex optimization framework via linear matrix inequalities (LMIs) will be considered. The benchmark problem is taken as an example to show how special controller gain variations can affect the performance of the closed-loop system.