Diagnosability of Discrete Event Systems under Sensor Attacks*

This paper is about the fault diagnosis of discrete event systems (DES) under time constraints. The DES contains a series of tasks with well-defined time intervals for each, collected during repetitive runs of the system. The repetitive runs allowed us to calculate average function time for each task. The objective is to propose an algorithm that verifies the correct behavior of each individual task within the system. To illustrate the principal of the algorithm, we will apply it with two concrete examples. This algorithm is easy to implement and operate, and is very useful for all types of systems.

The problem of safe failure diagnosis in discrete event systems is addressed. Starting from the standard definition of diagnosability of discrete event systems, which deals with the problem of detecting the occurrence of an unobservable event using the available observations on the system, the problem of performing the detection before the system executes a forbidden string is introduced. For example, this constraint could be required to prevent local faults from developing into failures that could cause safety hazards. This idea results in a new language property for discrete event systems called safe diagnosability. Necessary and sufficient conditions to test this language property are presented. Moreover, the problem of explicitly taking into account safe diagnosability as a requirement in system design is discussed.

Recently, safe diagnosability of discrete event systems (DESs) was investigated by Paoli and Lafortune, which was viewed as the first necessary step of fault-tolerant supervision. In this paper, we consider the problem of safe diagnosability in the framework of stochastic discrete event systems (SDESs). We define the notion of safe diagnosability for stochastic automata, in which fault detection occurs before any given forbidden string in the failed mode of system is executed. The relationship between diagnosability and safe diagnosability for SDESs is analyzed. In particular, a necessary and sufficient condition for safe diagnosability of SDESs is presented by constructing the recognizer of illegal language and the safe diagnoser. Some examples are described to illustrate the results.

In this paper a new approach to fault diagnosis of discrete events systems modeled with Petri nets is presented. Fault events are associated to a subset of unobservable transitions of the net. Negative markings, called generalized marking, are used to estimate the explanations of the observed transitions. The generalized markings allow to cope with the problems related with the explosion of the state space estimation in a very efficient way in terms of memory usage. This feature makes the proposed approach suitable for the on-line fault diagnosis of discrete event systems.

formula omitted]-diagnosability for active on-line diagnosis in discrete event systems

Towards low-cost fault diagnosis in large component-based systems

Introduction to Discrete Event Systems is a comprehensive introduction to the field of discrete event systems, offering a breadth of coverage that makes the material accessible to readers of varied backgrounds. The book emphasizes a unified modeling framework that transcends specific application areas, linking the following topics in a coherent manner: language and automata theory, supervisory control, Petri net theory, Markov chains and queuing theory, discrete-event simulation, and concurrent estimation techniques. This edition includes recent research results pertaining to the diagnosis of discrete event systems, decentralized supervisory control, and interval-based timed automata and hybrid automata models.

The paper deals with the definition of procedure that enables one to determine, for a given plant, if all faults can be detected and located after a finite sequence of observable events. More formally, the diagnosability is the property that every fault can be correctly detected from the observable events of the system after its occurrence no later than a bounded number of events. In this paper, the diagnosability problem of Discrete Event Systems (DESs) is studied. As modeling tool, finite-state automaton in an event-based framework is used. A necessary and sufficient condition of diagnosability of such systems is proposed. The results proposed in this paper allow checking the diagnosability of discrete event systems in an efficient way, i.e. in polynomial time.

In the conventional framework for failure diagnosis of partially observed discrete event systems, it is assumed that, for each event, the corresponding output symbol is determined uniquely. However, this assumption does not hold in discrete event systems such as a mobile system, where an output symbol depends on not only an event but also a state at which the event occurs. In this paper, we model such discrete event systems by Mealy automata with nondeterministic output functions, and consider a decentralized failure diagnosis problem that requires any occurrence of a failure should be detected by at least one local diagnoser within a uniformly bounded number of steps. We present a necessary and sufficient condition for the existence of a decentralized diagnoser. We then present an algorithm for verifying this condition. Moreover, we present a method for synthesizing local online diagnosers.

Diagnosability is an important property in the field of fault diagnosis. In this paper, a novel approach based on logical formula is proposed to verify diagnosability of Discrete event systems(DESs). CNFFSM is defined to represent a new model for DES. Each transition in DES can be described as a clause. According to CNF-FSM, we construct a CNF-diagnoser. Based on the resolution principle and CNF-diagnoser, an algorithm is presented to test whether the failure events can be detected or not in a finite number of observable events. Our algorithm can be applied in both off-line diagnosis and on-line diagnosis. Experimental results show that our algorithm can solve the diagnosability problem efficiently.

A framework is introduced for fault diagnosis in timed discrete-event systems. In this approach, the required estimates for system condition are updated only when the output changes or when deadlines associated with output changes expire. Thus updates at every clock tick are not required. This in many cases results in reduction in online computing requirements and in the size of the diagnosis system, at the expense of more offline design calculations. The issue of failure diagnosability is also discussed.

In this paper, we review some recent results about the use of dynamic observers for fault diagnosis of discrete event systems. Dynamic observers can switch sensors on or off, thus dynamically changing the set of events they wish to observe. We study the dynamic diagnoser synthesis problem and some related optimization problems.

This paper addresses the problem of online fault detection and diagnosis in discrete event systems modeled by labeled Petri nets and using Integer Linear Programming Problem (ILPP) solutions. In particular, unobservable (silent) transitions model faults and both observable and unobservable transitions model the nominal system behavior. Furthermore, observable transitions exhibit a kind of non determinism since several different transitions may share the same event label. This paper proposes two diagnosers that work in two different system settings. The first one is a centralized fault detection strategy: the diagnoser waits for an observable event and an algorithm defines and solves some ILPPs to decide whether the system behavior is normal or may exhibit some faults. In the second setting, the system consists of a set of interacting PN modules and each module is monitored by a diagnoser that has local information on the module structure. Moreover, each diagnoser observes and detects the faults of the module it is attached to and shares information in some of its places that are shared with other modules of the system. Some case studies show the two different approaches and point out the peculiarities of the proposed strategies.

The problem of fault diagnosis has become an active research topic in Discrete Event Systems, which can be applied into many different areas, such as communication networks, process control, flexible manufacturing systems, software engineering, etc. This brief explores the effect of communication delays on the event-based fault diagnosis in networked discrete event systems (NDESs), which can be modeled by finite state automata. Firstly, the framework of NDESs is introduced, and a novel concept of networked diagnosability ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$Net$ </tex-math></inline-formula> -diagnosability) is formally defined. Subsequently, the necessary and sufficient conditions for the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$Net$ </tex-math></inline-formula> -diagnosability in the partially observable outputs are provided, and the corresponding <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$Net$ </tex-math></inline-formula> -diagnosability can be verified by an effective diagnoser, which can be obtained by a constructed algorithm. Finally, an illustrative example is given to demonstrate the proposed method. The current results about networked diagnosability will be conducive to maintain the information security in the actual cyber-physical systems.

This paper deals with fault diagnosis and fault prognosis in discrete event systems described by sequences of events. We are interested in large industrial systems where the modelling with tools usually used for discrete event system (automata, Petri net…) is a complex task. The description of DES is made with a statistical model composed of events recorded from the considered DES and regrouped in the form of temporal windows. In the first phase, the theoretical framework is developed to perform diagnosis and prognosis using statistical model and temporal window concept. The second phase uses the result of the first phase to train radial basic function neural networks that preform online diagnosis and online prognosis. A case study and several examples serving to illuminate the developed approach are given.