Lilian Kawakami Carvalho

Robust diagnosis of discrete event systems against intermittent loss of observations

In the usual approaches to fault diagnosis of discrete event systems it is assumed that not only all sensors work properly but also all information reported by sensors always reaches the diagnoser. Any bad sensor operation or communication failure between sensors and the diagnoser can be regarded as loss of observations of events initially assumed as observable. In such situations, it may be possible that either the diagnoser stands still or report some wrong information regarding the fault occurrence. In this paper we assume that intermittent loss of observations may occur and we propose an automaton model based on a new language operation (language dilation) that takes it into account. We refer to this problem as robust diagnosability against intermittent loss of observations (or simply robust diagnosability, where the context allows). We present a necessary and sufficient condition for robust diagnosability in terms of the language generated by the original automaton and propose two tests for robust language diagnosability, one that deploys diagnosers and another one that uses verifiers. We also extend the results to robust codiagnosability against intermittent loss of observations.

Brief paperRobust diagnosis of discrete-event systems against permanent loss of observations☆

We consider the problem of diagnosing the occurrence of a certain unobservable event of interest, the fault event, in the operation of a partially-observed discrete-event system subject to permanent loss of observations modeled by a finite-state automaton. Specifically, it is assumed that certain sensors for events that would a priori be observable may fail at the outset, thereby resulting in a loss of observable events; the diagnostic engine is not directly aware of such sensor failures. We explore a previous definition of robust diagnosability of a given fault event despite the possibility of permanent (and unknown a priori) loss of observations and present a polynomial time verification algorithm to verify robust diagnosability and a methodology to perform online diagnosis in this scenario using a set of partial diagnosers.

Generalized Robust Diagnosability of Discrete Event Systems

Abstract We address the problem of robust diagnosability of discrete event systems described by a class of automata, where each automaton in the class generates a distinct language. We introduce a new definition which generalizes all previous definitions of robust diagnosability; for this reason it is referred here to as generalized robust diagnosability. We also present a necessary and sufficient condition for the generalized robust diagnosability and propose a polynomial time algorithm for its verification.

Robust diagnosability of discrete event systems subject to intermittent sensor failures

The modeling of physical systems using discrete event models assumes that a set of sensors always report the event occurrences correctly. However, bad sensor operation can result in loss of observability of the events associated with the malfunctioning sensors. If one or more sensors fail, it may be possible that either the diagnoser stands still or provide wrong information on the fault occurrence. This paper assumes that intermittent sensor failures may occur and deals with the problem of fault diagnosis in the presence of intermittent sensor failures. To this end, an automaton model for intermittent sensor failures based on a newly proposed language operation (language dilation) is presented in the paper. Necessary and sufficient conditions for robust diagnosability against intermittent sensor failure are also given in the paper. The development of a robust diagnoser that copes with intermittent sensor failures is another contribution of the paper.

Robust Supervisory Control Against Intermittent Loss of Observations

Abstract We address in this paper the design of robust supervisors for discrete-event systems subject to intermittent loss of observations. We present two definitions of robust observability: a more restrictive one that requires that the language achieved by the supervisor that control the nominal plant be also achieved by the robust supervisor, and a weaker one that also takes into account possible observation of the events that are subject of intermittent loss of observations. Necessary and sufficient conditions for the existence of robust supervisors that make the controlled system achieve weakly and strongly robust observable languages are also presented. A running example illustrates all the results presented in the paper.

Detection and prevention of actuator enablement attacks in supervisory control systems

The deployment of control systems with network-connected components nowadays has made feedback control systems vulnerable to attacks over the network. This paper considers the problem of intrusion detection and prevention in supervisory control systems, where the attacker has the ability to enable vulnerable actuator events that are disabled by the supervisor. We present a mathematical model for the system under such actuator enablement attacks and propose a defense strategy that detects attacks online and disables all controllable events after an attack is detected. We develop an algorithm for verifying whether the system can prevent damage from attacks with the proposed defense strategy, where damage is modeled as the reachability of a pre-defined set of “unsafe” states. The technical condition of interest that is necessary and sufficient in this context is characterized; it is termed “AE-safe controllability”. Finally, we illustrate the methodology with a traffic system example.

DESLAB: A scientific computing program for analysis and synthesis of discrete-event systems

Abstract In this paper we present deslab, a scientific computing program written in python , for the development of algorithms for analysis and synthesis of discrete event systems (DES) modeled as automata. The main objective of deslab is to provide a unified tool that integrates automata, graph algorithms, and numerical calculations. deslab also allows the definition of symbolic variables of type automaton and incorporates concise instructions to manipulate, operate, analyze and visualize these variables, with a syntax and an abstraction level close to the notation used in DES theory. Using the proposed set of instructions and basic control structures of python language, deslab can be easily extended, giving rise to new functions and toolboxes, according to the users’ needs.

New algorithms for verification of relative observability and computation of supremal relatively observable sublanguage

In this paper, we present a new property of relative observability, and based on this property, we propose two algorithms to deal with relative observability of regular languages. The first algorithm, that has polynomial complexity, verifies if a regular language is relatively observable. The second algorithm, that computes the supremal relatively observable sublanguage of a given regular language, has exponential complexity in the number of the states of the automaton that marks the specification language, being, therefore, considerably smaller than that of a recently proposed algorithm, which has doubly exponential complexity.

Diagnosability of intermittent sensor faults in discrete event systems

We address in this paper the problem of diagnosing intermittent sensor faults. In order to do so, we modify the model of intermittent loss of observations proposed in the literature to account for sensor malfunction only. Using this model together with a modified label automaton, it will be possible to change the problem of detecting intermittent sensor faults into a problem of diagnosing a language generated by an automaton in the presence of intermittent faults, where the fault event will be an unobservable event that models the non-detection of the event to be registered by the sensor under consideration. In this regard, we present necessary and sufficient conditions for diagnosability of intermittent sensor faults and propose a test based on diagnoser automaton to verify intermittent sensor fault diagnosability.

Supervisory Control-Based Navigation Architecture: A New Framework for Autonomous Robots in Industry 4.0 Environments

Industry 4.0 is characterized by an increasing dependence on automation and interconnection of systems due to the need for more efficient, autonomous, and customizable processes, and so, mobile robot navigation becomes an important tool. In this paper, we present a general methodology for mobile robot navigation in industrial environments in which the open-loop behavior of the robot and the specifications are based on automata. We build a modular supervisor, which is the conjunction of two supervisors: the first one that enforces the robot to follow the path defined by a planner and the second one that guarantees the satisfaction of the specifications such as prevention of collisions and task and movement management. The proposed navigation architecture allows decentralized implementation, in which the modular supervisor is embedded in the mobile robot, whereas the planner runs in an external agent. Such a feature makes the adaptation of the proposed navigation architecture to different environments easy. The navigation architecture proposed in this paper is illustrated by means of a simulation in a hypothetical environment that resembles a smart factory.