No abstract

No abstract

No abstract

In this research, we utilize semantic technology for robust early diagnosis and decision support. We present a lightweight platform that provides the enduser with direct access to the data through an ontology, and enables detection of any forthcoming faults by considering the data only from the reliable sensors. Concurrently, it indicates the actual sources of the detected faults, enabling mitigation action to be taken. Our work is focused on systems that require only real-time data and a restricted part of the historic data, such as fuel cell stack systems. First, we present an upper-level ontology that captures the semantics of such monitored systems and then we present the structure of the platform. Next, we specialize on the fuel cell paradigm and we provide a detailed description of our platform's functionality that can aid future servicing problem reporting applications.

In this paper, a novel model for dynamic reliability analysis of a PEM fuel cell system is developed using Modelica language in order to account for multi-state dynamics and aging. The modelling approach constitutes the combination of physical and stochastic sub-models with shared variables. The physical model consist of deterministic calculations of the system state described by variables such as temperature, pressure, mass flow rates and voltage output. Additionally, estimated component degradation rates are also taken into account.The non-deterministic model, on the other hand, is implemented with stochastic Petri nets which represent different events that can occur at random times during fuel cell lifetime. A case study of effects of a cooling system on fuel cell performance was investigated. Monte Carlo simulations of the process resulted in a distribution of system parameters, thus providing an estimate of best and worst scenarios of a fuel cell lifetime.

No abstract

As a possible alternative energy source, hydrogen fuel cells, especially Polymer Electrolyte Membrane (PEM) fuel cells, have received much more attention in the last few decades, which have already been equipped in many applications. A series of studies have been devoted to PEM fuel cell fault diagnosis to ensure its reliability during its lifetime, but due to the complexity of PEM fuel cell systems and incomplete PEM fuel cell test protocols, it is difficult to test various PEM fuel cell failure modes, thus the performance of fault diagnostic techniques cannot be fully investigated. On this basis, it is necessary to develop a reliable PEM fuel cell model with capability of simulating various PEM fuel cell faults. In this study, a hybrid model is developed to represent the behavior of PEM fuel cells in both continuous and discrete-time domains. With a continuous-time domain sub-model, various aspects of PEM fuel cell behavior can be simulated, including fluid, thermal, and electro-chemical dynamics. Moreover, the PEM fuel cell failure modes are implemented with stochastic Petri nets in the discrete-time domain. Based on the developed hybrid model, various PEM fuel cell failure modes can be simulated and their effects on the system performance can be observed. With the simulated data under different conditions, the performance of fault diagnostic techniques can be better evaluated by studying their performance in different failure mode scenarios.

No abstract