Strong detectability and observers

Conditions are given for the existence of observers that estimate unmeasured outputs on the basis of partial information on the input and the state. The concepts of strong detectability and strong observability, introduced before in the literature for discrete-time systems only, are defined and studied for continuous-time systems. It is shown that there are two different concepts of strong detectability, which coincide for discrete-time systems. Algebraic conditions for either concept are given. It is shown that these concepts are intimately related to the existence of strong observers, i.e. observers that only use the output of the plant.     

On unified concepts of detectability and observability for continuous-time stochastic systems

This paper is concerned with detectability and observability of continuous-time stochastic linear systems. By adopting the idea used in defining these two concepts for time-varying systems and Markovian jump linear systems that have been studied in the literature, corresponding definitions for continuous-time stochastic linear systems are proposed. These two definitions are not only able to unify some recent definitions on these two concept reported in the literature, but also allow us to propose an efficient rank criterion to test observability of continuous-time stochastic linear systems. It seems that this rank criterion is quite analogous to the rank criterion for deterministic linear systems. With the help of these two concepts and the new criteria, the stochastic Lyapunov equation is revisited and some recent published work on this equation are generalized. Numerical examples are given to illustrate the effectiveness of the proposed approach.     

Observability of hybrid discrete-continuous systems

We study the statement and solvability of observability problems in linear stationary hybrid discrete-continuous dynamical systems. Necessary and sufficient observability conditions expressed directly via the system parameters are derived. We consider linear observability problems and the dual controllability and reachability problems. The problem of computing the minimum number of inputs for which the system has a given observability is discussed. An example illustrating the results is presented.     

On the definiteness of quadratic functionals

In this paper we derive various characterizations of strong observability (or observability with unknown inputs) of time-dependent linear systems

Robust state estimation and fault diagnosis for uncertain hybrid systems

Robust state estimation and fault diagnosis are challenging problems in the research into hybrid systems. In this paper a novel robust hybrid observer is proposed for a class of hybrid systems with unknown inputs and faults. Model uncertainties, disturbances and faults are represented as structured unknown inputs. The robust hybrid observer consists of a mode observer for mode identification and a continuous observer for continuous state estimation and mode transition detection. It is shown that the mode can be identified correctly and the continuous state estimation error is exponentially uniformly bounded. Robustness to model uncertainties and disturbances can be guaranteed for the hybrid observer by disturbance decoupling. Furthermore, the detectability and mode identifiability conditions are rigorously analyzed. On the basis of the robust hybrid observer, a robust fault detection and isolation scheme is presented also in the paper. Simulations of a hybrid four-tank system show the proposed approach is effective.     

Optimal control of linear stochastic evolution equations in Hilbert spaces and uniform observability

In this paper we study the existence of the optimal (minimizing) control for a tracking problem, as well as a quadratic cost problem subject to linear stochastic evolution equations with unbounded coefficients in the drift. The backward differential Riccati equation (BDRE) associated with these problems (see [2], for finite dimensional stochastic equations or [21], for infinite dimensional equations with bounded coefficients) is in general different from the conventional BDRE (see [10], [18]). Under stabilizability and uniform observability conditions and assuming that the control weight-costs are uniformly positive, we establish that BDRE has a unique, uniformly positive, bounded on ℝ ₊ and stabilizing solution. Using this result we find the optimal control and the optimal cost. It is known [18] that uniform observability does not imply detectability and consequently our results are different from those obtained under detectability conditions (see [10]).      

Controllability and observability of coordinated linear systems

This paper develops concepts of controllability and observability for the class of coordinated linear systems, a special class of linear hierarchical systems with a top-to-bottom information structure. The usual concepts of reachability and indistinguishability for linear systems are refined, taking into account the decentralized nature of coordinated linear systems. The corresponding controllability and observability decompositions are derived, and several concepts of controllability and observability for coordinated linear systems are introduced and discussed. These results are then combined to a definition and characterization of stabilizability via dynamic measurement feedback for this class of systems.     

Linear stationary control systems over the Boolean semiring and their graphs of modules

We consider linear stationary dynamical systems over the Boolean semiring. We analyze the properties of complete observability, identifiability, attainability, and controllability of a system. We define the notion of the “graph of modules” of totally controllable totally attainable Boolean linear stationary systems by analogy with spaces of modules in the case of systems over fields. The above-mentioned graph is described in the simplest case of one-dimensional inputs and outputs. We prove the weak connectedness of this oriented graph.     

Controllability and Observability of Linear Quaternion-valued Systems

The aim of this paper is to define an extension of the controllability and observability for linear quaternion-valued systems (QVS). Some criteria for controllability and observability are derived, and the minimum norm control and duality theorem are also investigated. Compared with real-valued or complex-valued linear systems, it is shown that the classical Caylay-Hamilton Theorem as well as Popov-Belevitch-Hautus (PBH) type controllability and observability test do not hold for linear QVS. Hence, a modified PBH type necessary condition is studied for the controllability and observability, respectively. Finally, some examples are given to illustrate the effectiveness of the obtained results.     

On the observability of linear,time-invariant systems with unknown inputs

This paper considers the observation of linear, time-invariant dynamical systems in the general case in which some of the input functions are unknown. By arguments based on the concepts of controlled and conditioned invariance, a convenient expression for the observability subspace is found which includes the well-known expression for the case in which the input functions are given.     

Some sufficient conditions for the observability of nonlinear systems

Some sufficient conditions are presented for the observability of systems described by nonlinear, ordinary differential equations with linear observations. The point which the authors especially emphasize is the fact that the results obtained are global in some sense. In the beginning, the observability problem is reduced to the uniqueness problem of solutions of some nonlinear integral equations for general nonlinear systems with linear observations. Then, for some restricted systems, a simple sufficient condition is derived. The relation between global and local observability for nonlinear systems is also considered.The authors wish to thank their colleagues in Nagoya University.     

Methods for constructing observers for linear dynamical systems under uncertainty

We consider various methods for constructing asymptotic observers for linear stationary MIMO systems under an unknown external perturbation. Square systems (in which the number of measured outputs coincides with that of unknown inputs) are considered separately. For such systems we propose a method for constructing observers that yields either an asymptotic observer (for systems of maximal relative order) or an arbitrarily accurate observer. Hyperoutput systems (in which the number of measured outputs exceeds that of unknown inputs) are considered separately as well. For such systems we propose a method for synthesizing asymptotic observers of the phase vector.     

Controllability and observability for a class of time-varying impulsive systems

This paper is concerned with the controllability and observability for a class of piecewise linear time-varying impulsive systems. Several sufficient and necessary conditions for state controllability and observability of such systems are established. Meanwhile, corresponding criteria for time-invariant impulsive systems are also obtained and the criteria are compared with the existing results.     

Conditions for distinguishability and observability of switched linear systems

A characterization for weak distinguishability of two linear controlled systems are derived. In addition, the observability of a special switched linear controlled system is considered, and necessary and sufficient conditions for observability are obtained.     

Complements to full order observer design for linear systems with unknown inputs

This note completes the results obtained in [M. Darouach, M. Zasadzinski, S.J. Xu, Full-order observers for linear systems with unknown inputs, IEEE Trans. Automat. Control., AC-39 (3) (1994) 606–609]. It presents a full order observer for linear systems with unknown inputs in the state and in the measurement equations. It gives a more general approach for the observer’s design than that presented in the above mentioned reference and it shows that all the obtained results are independent of the choice of the generalized inverses considered in the observer’s design. Continuous- and discrete-time systems are considered.     

Observability of nonlinear systems

New necessary and sufficient conditions are presented for the observability of systems described by nonlinear ordinary differential equations with nonlinear observations. The conditions are based on extension of the necessary and sufficient conditions for observability of time-varying linear systems to the linearized trajectory of the nonlinear system. The result is that the local observability of any initial condition can be readily determined, and the observability of the entire initial domain can be computed. The observability of constant parameters appearing in the differential equations is also considered. Examples are presented to illustrate the theory.This research was supported by the National Science Foundation, Grant No. NSF GK-10136.     

Observability of continuous Petri nets with infinite server semantics

Timed continuous Petri net (contPN) systems with infinite server semantics are nonlinear systems, particularly a subclass of piecewise linear (PWL) systems. This paper addresses several problems regarding the state observability of these systems. We assume that the initial marking/state is not known and by measuring the marking of some places we want to estimate all the others. First, a study of the different linear systems corresponding to a continuous Petri net system is performed. It is shown that in some cases, some of them are redundant, and so can be disregarded. The notion of distinguishable modes is introduced which helps us in giving a necessary and sufficient criterion for the observability in infinitesimal time. Structural observability, i.e., observability for all possible values of firing rates of transitions, is studied and it is proved that in some cases it can be reduced to a linear problem, even if the system is nonlinear. Using results from linear structured systems, the concept of weak structural or generic observability is considered.     

Generalized Gram Matrix and Its Application to the Observability Problem for Linear Nonsteady Systems

Sufficient conditions for the nondegeneracy of a generalized Gram matrix are obtained. In particular, it is shown that the generalized Gram matrix is nondegenerate for the Chebyshev systems of functions. An application of the results to the observability problems for linear nonsteady systems of ordinary differential equations are given. In terms of the observability matrix, necessary and sufficient conditions of the complete and total observability by means of finite-parameter solving operations are established.     

Unknown-input present-state observability of discrete-time linear systems

A generalized definition of the present-state unknown-input observability is given as well as the corresponding unobservable subspace for discrete-time linear dynamical systems whose initial state is known, unknown, or partially known. It is shown that the time sequence of the unobservable subspaces have interesting properties explicable via generalized reachability subspaces, namely, the properties of conditional monotonicity and those associated with it. Furthermore, an observer is constructed to determine the present state.