Dimitri Lefebvre

Fault Diagnosis and Prognosis With Partially Observed Petri Nets

This paper concerns the prevention of faults in discrete event systems modeled with partially observed Petri nets (POPNs) that include the definition of sensors used to measure the events and markings. Observation sequences result from this modeling, and the firing sequences and initial marking consistent with a given observation sequence are systematically obtained. The degree of confidence of past and future states and events are computed: state estimation fault diagnosis and fault prediction result from this computation. Finally, diagnosability, detectability, and predictability are defined for observation sequences and POPNs and are discussed with respect to the sensor configuration.

Multimodel for the coupling of several dc/dc power converters on a dc bus

This paper concerns multisource renewable energy systems. It describes the design of an average state space model that brings a detailed physical explanation of the coupling and uncoupling of several dc/dc power converters on a dc bus. A multimodel is proposed for that purpose, which consists in a mathematical generic expression whose parameters change according to the dc/dc power converters coupled on the dc bus. This multimodel can be used as support for the design of hybrid control laws in order to optimize the energies transfers, according to the sources power variations and the load characteristics. It takes into account conduction losses in dc/dc power converters and makes it possible to evaluate the efficiency of the equipment.

Fault diagnosis and prognosis with partially observed stochastic Petri nets

This article concerns fault diagnosis and prognosis for stochastic discrete event systems. For this purpose, partially observed stochastic Petri nets are introduced that include the sensors used to measure events and markings and the Markovian stochastic dynamics used to represent failure processes. Timed observation sequences result from this modeling, and the probabilities of timed and untimed marking trajectories consistent with a given timed observation sequence are systematically computed. Diagnosis in terms of fault probability is obtained as a consequence and compared with the belief of faults that is usually used for diagnosis issues. Confidence factors based on fault probabilities are also proposed. Finally, state estimation and fault prediction are investigated, and probability of future faults is obtained as a consequence. An application case is studied to illustrate the method.

Petri nets control design for hybrid electrical energy systems

This paper describes a supervisory control strategy based on Petri nets for electrical energy transfers in multisource renewable energy systems. The aim is to optimize the energy transfers, according to the sources power variations and the load characteristics. For this purpose, the proposed Petri net controller calculates the operating mode of the multisource renewable energy system. This high level controller is combined with a low level one that tunes the power ratio provided by each source according to sources power (climatic conditions) and load variations. An estimation of the duty cycle value of the dc/dc power converters is used to fire the Petri nets transitions, to switch the power sources and also to drive the power ratios.

Early FDI based on residuals design according to the analysis of models of faults: application to DAMADICS

The increased complexity of plants and the development of sophisticated control systems have encouraged the parallel development of efficient rapid fault detection and isolation (FDI) systems. FDI in industrial system has lately become of great significance. This paper proposes a new technique for short time fault detection and diagnosis in nonlinear dynamic systems with multi inputs and multi outputs. The main contribution of this paper is to develop a FDI schema according to reference models of fault-free and faulty behaviors designed with neural networks. Fault detection is obtained according to residuals that result from the comparison of measured signals with the outputs of the fault free reference model. Then, Euclidean distance from the outputs of models of faults to themeasurements leads to fault isolation. The advantage of this method is to provide not only early detection but also early diagnosis thanks to the parallel computation of the models of faults and to the proposed decision algorithm. The effectiveness of this approach is illustrated with simulations on DAMADICS benchmark.

Deadlock-free scheduling for timed Petri net models combined with MPC and backtracking

This paper proposes algorithms that incrementally compute control sequences driving the marking of timed Petri nets from an initial value to a reference one, avoiding forbidden and dangerous states with a duration that approaches the minimal value. The proposed algorithms are applicable to a large class of discrete event systems in particular in the domain of flexible manufacturing, communication and computer science or transportation and traffic. To overcome the most burdensome part of the computations, only the sequences encoded in a small area of the reachability graph are worked out. The main contributions are to propose an estimation of the minimal duration of the remaining sequences to the reference based on the computation of the firing count vectors and a progressive search of the forbidden and dangerous states according to a backtracking phase. The approach is suitable for deadlock-free scheduling problems.

Faults prognosis using partially observed stochastic Petri nets

This article deals with the problem of fault prognosis in stochastic discrete event systems. For that purpose, partially observed stochastic Petri nets are considered to model the system with its sensors. The model represents both healthy and faulty behaviors of the system. Marking trajectories which are consistent with the measurements issued from the sensors are first obtained. Based on the events dates, the probabilities of the consistent trajectories are evaluated and a state estimation is obtained as a consequence. From the set of possible current states and their probabilities, a method to evaluate the probability of future fault is developed using a probabilistic model. An example is presented to illustrate the results.

Feasibility of piecewise-constant control sequences for timed continuous Petri nets

In this study, the determination of control actions for timed continuous Petri nets is investigated by the characterisation of attractive regions in marking space. In particular, attraction in finite time, which is important for practical issues, is considered. Based on the characterisation of attractive regions, the domain of admissible piecewise constant control actions is computed, and sufficient conditions to verify the feasibility of the control objectives are proposed. As a consequence, an iterative procedure is presented to compute piecewise constant control actions that correspond to local minimum time control for timed continuous Petri nets.

Estimation of the fault occurrence dates in DESs with partially observed stochastic Petri nets

Abstract In this paper, a method for fault occurrence date evaluation in stochastic discrete event systems is proposed. The system is modeled using partially observed stochastic Petri nets. This model includes the faulty behavior that represents the failure processes of the system. From timed observation sequences resulting from this modeling, our approach consists, first, on evaluating the probabilities of marking trajectories consistent with the observations. Diagnosis in terms of faults probability is obtained as a consequence. Thereafter, a probabilistic model based on the sum of random exponential variables is presented to determine the fault date probability.

Limitation of flow variation of timed continuous Petri Nets via Model Predictive Control

This paper addresses the problem of flow variation of timed continuous Petri Nets under infinite server semantics. Our goal is not only to demonstrate the impact of the weighting factor associated to the expression of the flow in the cost function, in reducing flow variation. It is also to find the optimal value and a control optimizing the defined cost function that drives the evolution of the system from an initial marking M0 to a desired one Md and reduces the high flow variation. The problem is studied using Model Predictive Control (MPC). The variation rate is reduced significantly whereas some oscillations around the desired marking arise. The optimal weighting factor is determined in order to limit the oscillations amplitude. Therefore, the flow is the smoothest and accordingly the excessive solicitation of actuators can be reduced.