This paper presents a neural network (NN) approach to detect and locate automatically multiple so... more This paper presents a neural network (NN) approach to detect and locate automatically multiple soft faults in complex wired networks using multi-sensor information fusion. The location process is based on monitoring the wired network topology by several sensors (reflectometers). The soft fault detection and location are achieved by Multi-Carrier Time Domain Reflectometry (MCTDR) combined with feedforward Multi-Layer Perceptron (MLP) neural network, trained by backpropagation algorithm. The NN ensures the data fusion between different reflectometers. The required datasets for training and testing the NN are generated by simulation of faults for various soft faults scenarios (fault locations and fault impedance). The effectiveness of the proposed approach is demonstrated by simulation for locating multiple soft faults in branched network.
Les recherches menées dans cette thèse portent sur le diagnostic de réseaux filaires complexes à ... more Les recherches menées dans cette thèse portent sur le diagnostic de réseaux filaires complexes à l’aide de la réflectométrie distribuée. L’objectif est de développer de nouvelles technologies de diagnostic en ligne, distribuées des réseaux complexes permettant la fusion de données ainsi que la communication entre les réflectomètres pour détecter, localiser et caractériser les défauts électriques (francs et non francs). Cette collaboration entre les réflectomètres permet de résoudre le problème d’ambiguïté de localisation des défauts et d’améliorer la qualité du diagnostic. La première contribution concerne la proposition d’une méthode basée sur la théorie des graphes permettant la combinaison de données entre les réflectomètres distribués afin de faciliter la localisation d’un défaut. L’amplitude du signal réfléchi est ensuite utilisée pour identifier le type du défaut et estimer son impédance. Cette estimation est basée sur la régénération du signal en compensant la dégradation sub...
Advances in Intelligent Systems and Computing, 2018
This contribution demonstrates the feasibility of integrating sensors communications on diagnosis... more This contribution demonstrates the feasibility of integrating sensors communications on diagnosis signal of wired networks. The objective of this study is to propose a new distributed diagnosis technology for complex wire networks to ensure the communication between sensors by using the transmitted part of the reflectometry signal. This communication contributes to improve the diagnosis quality by reflectometers data fusion. The MCTDR (Multi-carrier Time Domain Reflectometry) shows good performance (faults detection and localization) on distributed diagnosis with a precise control of the signals spectrum while avoiding interference. We use this signal as a carrier of information without degrading these good characteristics. This communication is robust to cable faults and noise. We thus exploit simultaneously the reflected part and the transmitted part of the signal.
This paper presents a neural network (NN) approach to detect and locate automatically multiple so... more This paper presents a neural network (NN) approach to detect and locate automatically multiple soft faults in complex wired networks using multi-sensor information fusion. The location process is based on monitoring the wired network topology by several sensors (reflectometers). The soft fault detection and location are achieved by Multi-Carrier Time Domain Reflectometry (MCTDR) combined with feedforward Multi-Layer Perceptron (MLP) neural network, trained by backpropagation algorithm. The NN ensures the data fusion between different reflectometers. The required datasets for training and testing the NN are generated by simulation of faults for various soft faults scenarios (fault locations and fault impedance). The effectiveness of the proposed approach is demonstrated by simulation for locating multiple soft faults in branched network.
IEEE Transactions on Electromagnetic Compatibility
This article presents the integration of sensor communication using the diagnosis signal of wired... more This article presents the integration of sensor communication using the diagnosis signal of wired networks. In order to cover the entire network, the diagnosis of branched networks requires the use of distributed diagnosis systems. The objective of this article is to propose a novel technique of ensuring efficient communication between distributed reflectometers (sensors) by using the transmitted part of the multicarrier time-domain reflectometry (MCTDR) signal. Our approach provides unambiguous fault location thanks to sensor data sharing and fusion, which improves the accuracy of fault location and the quality of the diagnosis. MCTDR has already been proven as an efficient method for online diagnosis by enabling precise bandwidth control while avoiding interferences. The main novelty of our technique is to inject an MCTDR signal carrying information which is capable of ensuring network diagnosis and reliable communication between several distributed sensors at the same time. This is done by exploiting, simultaneously, both the transmitted part and the reflected part of the MCTDR signal. The effectiveness of the proposed approach is evaluated by a series of simulations and experiments (based on field-programmable gate array implementation) on different types.
This paper presents a neural network (NN) approach to detect and locate automatically multiple so... more This paper presents a neural network (NN) approach to detect and locate automatically multiple soft faults in complex wired networks using multi-sensor information fusion. The location process is based on monitoring the wired network topology by several sensors (reflectometers). The soft fault detection and location are achieved by Multi-Carrier Time Domain Reflectometry (MCTDR) combined with feedforward Multi-Layer Perceptron (MLP) neural network, trained by backpropagation algorithm. The NN ensures the data fusion between different reflectometers. The required datasets for training and testing the NN are generated by simulation of faults for various soft faults scenarios (fault locations and fault impedance). The effectiveness of the proposed approach is demonstrated by simulation for locating multiple soft faults in branched network.
Les recherches menées dans cette thèse portent sur le diagnostic de réseaux filaires complexes à ... more Les recherches menées dans cette thèse portent sur le diagnostic de réseaux filaires complexes à l’aide de la réflectométrie distribuée. L’objectif est de développer de nouvelles technologies de diagnostic en ligne, distribuées des réseaux complexes permettant la fusion de données ainsi que la communication entre les réflectomètres pour détecter, localiser et caractériser les défauts électriques (francs et non francs). Cette collaboration entre les réflectomètres permet de résoudre le problème d’ambiguïté de localisation des défauts et d’améliorer la qualité du diagnostic. La première contribution concerne la proposition d’une méthode basée sur la théorie des graphes permettant la combinaison de données entre les réflectomètres distribués afin de faciliter la localisation d’un défaut. L’amplitude du signal réfléchi est ensuite utilisée pour identifier le type du défaut et estimer son impédance. Cette estimation est basée sur la régénération du signal en compensant la dégradation sub...
Advances in Intelligent Systems and Computing, 2018
This contribution demonstrates the feasibility of integrating sensors communications on diagnosis... more This contribution demonstrates the feasibility of integrating sensors communications on diagnosis signal of wired networks. The objective of this study is to propose a new distributed diagnosis technology for complex wire networks to ensure the communication between sensors by using the transmitted part of the reflectometry signal. This communication contributes to improve the diagnosis quality by reflectometers data fusion. The MCTDR (Multi-carrier Time Domain Reflectometry) shows good performance (faults detection and localization) on distributed diagnosis with a precise control of the signals spectrum while avoiding interference. We use this signal as a carrier of information without degrading these good characteristics. This communication is robust to cable faults and noise. We thus exploit simultaneously the reflected part and the transmitted part of the signal.
This paper presents a neural network (NN) approach to detect and locate automatically multiple so... more This paper presents a neural network (NN) approach to detect and locate automatically multiple soft faults in complex wired networks using multi-sensor information fusion. The location process is based on monitoring the wired network topology by several sensors (reflectometers). The soft fault detection and location are achieved by Multi-Carrier Time Domain Reflectometry (MCTDR) combined with feedforward Multi-Layer Perceptron (MLP) neural network, trained by backpropagation algorithm. The NN ensures the data fusion between different reflectometers. The required datasets for training and testing the NN are generated by simulation of faults for various soft faults scenarios (fault locations and fault impedance). The effectiveness of the proposed approach is demonstrated by simulation for locating multiple soft faults in branched network.
IEEE Transactions on Electromagnetic Compatibility
This article presents the integration of sensor communication using the diagnosis signal of wired... more This article presents the integration of sensor communication using the diagnosis signal of wired networks. In order to cover the entire network, the diagnosis of branched networks requires the use of distributed diagnosis systems. The objective of this article is to propose a novel technique of ensuring efficient communication between distributed reflectometers (sensors) by using the transmitted part of the multicarrier time-domain reflectometry (MCTDR) signal. Our approach provides unambiguous fault location thanks to sensor data sharing and fusion, which improves the accuracy of fault location and the quality of the diagnosis. MCTDR has already been proven as an efficient method for online diagnosis by enabling precise bandwidth control while avoiding interferences. The main novelty of our technique is to inject an MCTDR signal carrying information which is capable of ensuring network diagnosis and reliable communication between several distributed sensors at the same time. This is done by exploiting, simultaneously, both the transmitted part and the reflected part of the MCTDR signal. The effectiveness of the proposed approach is evaluated by a series of simulations and experiments (based on field-programmable gate array implementation) on different types.
Uploads
Papers by Ousama osman