Nonlinear dynamical systems identification and behavior prediction are difficult problems encount... more Nonlinear dynamical systems identification and behavior prediction are difficult problems encountered in many areas of industrial applications such as fault diagnosis and prognosis. In practice, the analytical description of a nonlinear system directly from observed data is a very challenging task because of the the too large number of the related parameters to be estimated. As a solution, multi-modeling approaches have lately been applied and consist in dividing the operating range of the system under study into different operating regions easier to describe by simpler functions to be combined. In order to take into consideration the uncertainty related to the available data as well as the uncertainty resulting from the nonlinearity of the system, evidence theory is of particular interest, because it permits the explicit modeling of doubt and ignorance. In the context of multi-modeling, information of doubt may be exploited to properly segment the data and take into account the uncertainty in the transitions between the operating regions. Recently, the Evidential Evolving Gustafson-Kessel algorithm (E2GK) has been proposed to ensure an online partitioning of the data into clusters that correspond to operating regions. Based on E2GK, a multi-modeling approach called E2GKpro is introduced in this paper, which dynamically performs the estimation of the local models by upgrading and modifying their parameters while data arrive. The proposed algorithm is tested on several datasets and compared to existing approaches. The results show that the use of virtual centroids in E2GKpro account for its robustness to noise and generating less operating regions while ensuring precise predictions.
The aim of this study is to analyze the Proton Exchange Membrane (PEM) water electrolysis. On the... more The aim of this study is to analyze the Proton Exchange Membrane (PEM) water electrolysis. On the basis of theoretical investigation the well know Bulter-Volmer equation and water transport characteristics through the solid electrolyte membrane were employed to simulate the electrode activation over potential and membrane ohmic over potential. Then, the simulation results were compared with the published experimental data.
The research activity in the PHM community is in full bloom and many efforts are being made to de... more The research activity in the PHM community is in full bloom and many efforts are being made to develop more realistic and reliable methodologies. However, there still exist very few real-world applications due to the complexity of the sys-tems of interest. Nonlinear dynamical systems identification and behavior prediction are difficult problems encountered in prognosis. The difficulty in switching from theory to practice can partially be explained by the existence of different kinds of uncertainty at each step of the implementation that must be taken into account with the appropriate tools. In this pa-per, we propose an evolving multi-modeling approach for the detection, the adaptation and the combination of local mod-els in order to analyze complex systems behavior. It relies on belief functions in order to take into consideration the uncer-tainty related to the available data describing the system as well as the uncertainty generated by the nonlinearity of the system. The informat...
2013 Conference on Control and Fault-Tolerant Systems (SysTol), 2013
ABSTRACT A wind farm is a complex system composed of several wind turbines operating in a non-sta... more ABSTRACT A wind farm is a complex system composed of several wind turbines operating in a non-stationary environment. Each of the wind turbines is subject to sudden and gradual faults due to operational and environmental conditions, aging etc. In order to assure an optimal power production and reduce maintenance costs, these faults have to be detected, isolated as soon as possible, and predicted. In this paper, an evolving classification method is proposed to achieve these requirements. The proposed approach is data-driven and does not require prior physical knowledge, in particular wind dynamics. It is based on the dynamic classification algorithm {\it AUDyC}. The considered features are determined according to the difference between generated electric powers regarding several operating modes. Normal operating modes are represented by classes in a decision space. Each new measure is classified on line. Indicators are computed to detect and isolate the occurrence of faults. Finally, a predictive method is implemented to forecast the degradation state of the wind turbine. A wind farm benchmark model, proposed for a fault diagnosis and fault tolerant control competition is used to highlight the efficiency of the proposed approaches.
Symbolic and Quantitative Approaches to Reasoning with Uncertainty, 2011
A new online clustering method, called E2GK (Evidential Evolving Gustafson-Kessel) is introduced ... more A new online clustering method, called E2GK (Evidential Evolving Gustafson-Kessel) is introduced in the theoretical framework of belief functions. The algorithm enables an online partitioning of data streams based on two existing and efficient algorithms: Evidantial c-Means (ECM) and Evolving Gustafson-Kessel (EGK). E2GK uses the concept of credal partition of ECM and adapts EGK, offering a better interpretation of the data structure. Experiments with synthetic data sets show good performances of the proposed algorithm compared to the original online procedure.
2011 IEEE Conference on Prognostics and Health Management, 2011
Diagnostics and prognostics of health states are important activities in the maintenance process ... more Diagnostics and prognostics of health states are important activities in the maintenance process strategy of dynamical systems. Many approaches have been developed for this purpose and we particularly focus on data-driven methods which are increasingly applied due to the availability of various cheap sensors. Most data-driven methods proposed in the literature rely on probability density estimation. However, when the training data are limited, the estimated parameters are no longer reliable. This is particularly true for data in faulty states which are generally expensive and difficult to obtain. In order to solve this problem, we propose to use the theory of belief functions as described by Dempster, Shafer (Theory of Evidence) and Smets (Transferable Belief Model). A few methods based on belief functions have been proposed for diagnostics and prognostics of dynamical systems. Among these methods, Evidential Hidden Markov Models (EvHMM) seems promising and extends usual HMM to belief functions. Inference tools in EvHMM have already been developed, but parameter training has not fully been considered until now or only with strong assumptions. In this paper, we propose to complete the generalization of HMM to belief functions with a method for automatic parameter training. The generalization of this training procedure to more general Time-Sliced Temporal Evidential Network (TSTEN) is discussed paving the way for a further generalization of Dynamic Bayesian Network to belief functions with potential applications to diagnostics and prognostics. An application to time series classification is proposed.
Nonlinear dynamical systems identification and behavior prediction are difficult problems encount... more Nonlinear dynamical systems identification and behavior prediction are difficult problems encountered in many areas of industrial applications such as fault diagnosis and prognosis. In practice, the analytical description of a nonlinear system directly from observed data is a very challenging task because of the the too large number of the related parameters to be estimated. As a solution, multi-modeling approaches have lately been applied and consist in dividing the operating range of the system under study into different operating regions easier to describe by simpler functions to be combined. In order to take into consideration the uncertainty related to the available data as well as the uncertainty resulting from the nonlinearity of the system, evidence theory is of particular interest, because it permits the explicit modeling of doubt and ignorance. In the context of multi-modeling, information of doubt may be exploited to properly segment the data and take into account the uncertainty in the transitions between the operating regions. Recently, the Evidential Evolving Gustafson-Kessel algorithm (E2GK) has been proposed to ensure an online partitioning of the data into clusters that correspond to operating regions. Based on E2GK, a multi-modeling approach called E2GKpro is introduced in this paper, which dynamically performs the estimation of the local models by upgrading and modifying their parameters while data arrive. The proposed algorithm is tested on several datasets and compared to existing approaches. The results show that the use of virtual centroids in E2GKpro account for its robustness to noise and generating less operating regions while ensuring precise predictions.
International Journal of Approximate Reasoning, 2012
A new online clustering method called E2GK (Evidential Evolving Gustafson-Kessel) is introduced. ... more A new online clustering method called E2GK (Evidential Evolving Gustafson-Kessel) is introduced. This partitional clustering algorithm is based on the concept of credal partition defined in the theoretical framework of belief functions. A credal partition is derived online by applying an algorithm resulting from the adaptation of the Evolving Gustafson-Kessel (EGK) algorithm. Online partitioning of data streams is then possible with a meaningful interpretation of the data structure. A comparative study with the original online procedure shows that E2GK outperforms EGK on different entry data sets. To show the performance of E2GK, several experiments have been conducted on synthetic data sets as well as on data collected from a real application problem. A study of parameters' sensitivity is also carried out and solutions are proposed to limit complexity issues.
The implementation of Sustainable Development (SD) within an Organization is a difficult task. Th... more The implementation of Sustainable Development (SD) within an Organization is a difficult task. This is due to the fact that it is difficult to deal with conflicting and incommensurable aspects such as environmental, economic and social dimensions. In this paper we have used a Multi-Criteria Decision Aid (MCDA) methodology to cope with these difficulties. MCDA methodology offers the opportunity to avoid monetary valuation of the different dimensions of the SD. These dimensions are not substitutable for one another and all have a role to play. There is an abundance of possible aggregation procedures in MCDA methodology. In this paper we have proposed an innovative method to choose a suitable aggregation procedure for SD problems. Real life case studies of the implementation of an outranking approach (i.e., ELECTRE) and of a mono-criterion synthesis approach (i.e., MAUT approaches based on the Choquet integral) were done to respectively rank 22 SD strategic actions within an expertise Institute and rank 20 practical operational actions to control energy consumption of the Institute's buildings.
Nonlinear dynamical systems identification and behavior prediction are difficult problems encount... more Nonlinear dynamical systems identification and behavior prediction are difficult problems encountered in many areas of industrial applications such as fault diagnosis and prognosis. In practice, the analytical description of a nonlinear system directly from observed data is a very challenging task because of the the too large number of the related parameters to be estimated. As a solution, multi-modeling approaches have lately been applied and consist in dividing the operating range of the system under study into different operating regions easier to describe by simpler functions to be combined. In order to take into consideration the uncertainty related to the available data as well as the uncertainty resulting from the nonlinearity of the system, evidence theory is of particular interest, because it permits the explicit modeling of doubt and ignorance. In the context of multi-modeling, information of doubt may be exploited to properly segment the data and take into account the uncertainty in the transitions between the operating regions. Recently, the Evidential Evolving Gustafson-Kessel algorithm (E2GK) has been proposed to ensure an online partitioning of the data into clusters that correspond to operating regions. Based on E2GK, a multi-modeling approach called E2GKpro is introduced in this paper, which dynamically performs the estimation of the local models by upgrading and modifying their parameters while data arrive. The proposed algorithm is tested on several datasets and compared to existing approaches. The results show that the use of virtual centroids in E2GKpro account for its robustness to noise and generating less operating regions while ensuring precise predictions.
The aim of this study is to analyze the Proton Exchange Membrane (PEM) water electrolysis. On the... more The aim of this study is to analyze the Proton Exchange Membrane (PEM) water electrolysis. On the basis of theoretical investigation the well know Bulter-Volmer equation and water transport characteristics through the solid electrolyte membrane were employed to simulate the electrode activation over potential and membrane ohmic over potential. Then, the simulation results were compared with the published experimental data.
The research activity in the PHM community is in full bloom and many efforts are being made to de... more The research activity in the PHM community is in full bloom and many efforts are being made to develop more realistic and reliable methodologies. However, there still exist very few real-world applications due to the complexity of the sys-tems of interest. Nonlinear dynamical systems identification and behavior prediction are difficult problems encountered in prognosis. The difficulty in switching from theory to practice can partially be explained by the existence of different kinds of uncertainty at each step of the implementation that must be taken into account with the appropriate tools. In this pa-per, we propose an evolving multi-modeling approach for the detection, the adaptation and the combination of local mod-els in order to analyze complex systems behavior. It relies on belief functions in order to take into consideration the uncer-tainty related to the available data describing the system as well as the uncertainty generated by the nonlinearity of the system. The informat...
2013 Conference on Control and Fault-Tolerant Systems (SysTol), 2013
ABSTRACT A wind farm is a complex system composed of several wind turbines operating in a non-sta... more ABSTRACT A wind farm is a complex system composed of several wind turbines operating in a non-stationary environment. Each of the wind turbines is subject to sudden and gradual faults due to operational and environmental conditions, aging etc. In order to assure an optimal power production and reduce maintenance costs, these faults have to be detected, isolated as soon as possible, and predicted. In this paper, an evolving classification method is proposed to achieve these requirements. The proposed approach is data-driven and does not require prior physical knowledge, in particular wind dynamics. It is based on the dynamic classification algorithm {\it AUDyC}. The considered features are determined according to the difference between generated electric powers regarding several operating modes. Normal operating modes are represented by classes in a decision space. Each new measure is classified on line. Indicators are computed to detect and isolate the occurrence of faults. Finally, a predictive method is implemented to forecast the degradation state of the wind turbine. A wind farm benchmark model, proposed for a fault diagnosis and fault tolerant control competition is used to highlight the efficiency of the proposed approaches.
Symbolic and Quantitative Approaches to Reasoning with Uncertainty, 2011
A new online clustering method, called E2GK (Evidential Evolving Gustafson-Kessel) is introduced ... more A new online clustering method, called E2GK (Evidential Evolving Gustafson-Kessel) is introduced in the theoretical framework of belief functions. The algorithm enables an online partitioning of data streams based on two existing and efficient algorithms: Evidantial c-Means (ECM) and Evolving Gustafson-Kessel (EGK). E2GK uses the concept of credal partition of ECM and adapts EGK, offering a better interpretation of the data structure. Experiments with synthetic data sets show good performances of the proposed algorithm compared to the original online procedure.
2011 IEEE Conference on Prognostics and Health Management, 2011
Diagnostics and prognostics of health states are important activities in the maintenance process ... more Diagnostics and prognostics of health states are important activities in the maintenance process strategy of dynamical systems. Many approaches have been developed for this purpose and we particularly focus on data-driven methods which are increasingly applied due to the availability of various cheap sensors. Most data-driven methods proposed in the literature rely on probability density estimation. However, when the training data are limited, the estimated parameters are no longer reliable. This is particularly true for data in faulty states which are generally expensive and difficult to obtain. In order to solve this problem, we propose to use the theory of belief functions as described by Dempster, Shafer (Theory of Evidence) and Smets (Transferable Belief Model). A few methods based on belief functions have been proposed for diagnostics and prognostics of dynamical systems. Among these methods, Evidential Hidden Markov Models (EvHMM) seems promising and extends usual HMM to belief functions. Inference tools in EvHMM have already been developed, but parameter training has not fully been considered until now or only with strong assumptions. In this paper, we propose to complete the generalization of HMM to belief functions with a method for automatic parameter training. The generalization of this training procedure to more general Time-Sliced Temporal Evidential Network (TSTEN) is discussed paving the way for a further generalization of Dynamic Bayesian Network to belief functions with potential applications to diagnostics and prognostics. An application to time series classification is proposed.
Nonlinear dynamical systems identification and behavior prediction are difficult problems encount... more Nonlinear dynamical systems identification and behavior prediction are difficult problems encountered in many areas of industrial applications such as fault diagnosis and prognosis. In practice, the analytical description of a nonlinear system directly from observed data is a very challenging task because of the the too large number of the related parameters to be estimated. As a solution, multi-modeling approaches have lately been applied and consist in dividing the operating range of the system under study into different operating regions easier to describe by simpler functions to be combined. In order to take into consideration the uncertainty related to the available data as well as the uncertainty resulting from the nonlinearity of the system, evidence theory is of particular interest, because it permits the explicit modeling of doubt and ignorance. In the context of multi-modeling, information of doubt may be exploited to properly segment the data and take into account the uncertainty in the transitions between the operating regions. Recently, the Evidential Evolving Gustafson-Kessel algorithm (E2GK) has been proposed to ensure an online partitioning of the data into clusters that correspond to operating regions. Based on E2GK, a multi-modeling approach called E2GKpro is introduced in this paper, which dynamically performs the estimation of the local models by upgrading and modifying their parameters while data arrive. The proposed algorithm is tested on several datasets and compared to existing approaches. The results show that the use of virtual centroids in E2GKpro account for its robustness to noise and generating less operating regions while ensuring precise predictions.
International Journal of Approximate Reasoning, 2012
A new online clustering method called E2GK (Evidential Evolving Gustafson-Kessel) is introduced. ... more A new online clustering method called E2GK (Evidential Evolving Gustafson-Kessel) is introduced. This partitional clustering algorithm is based on the concept of credal partition defined in the theoretical framework of belief functions. A credal partition is derived online by applying an algorithm resulting from the adaptation of the Evolving Gustafson-Kessel (EGK) algorithm. Online partitioning of data streams is then possible with a meaningful interpretation of the data structure. A comparative study with the original online procedure shows that E2GK outperforms EGK on different entry data sets. To show the performance of E2GK, several experiments have been conducted on synthetic data sets as well as on data collected from a real application problem. A study of parameters' sensitivity is also carried out and solutions are proposed to limit complexity issues.
The implementation of Sustainable Development (SD) within an Organization is a difficult task. Th... more The implementation of Sustainable Development (SD) within an Organization is a difficult task. This is due to the fact that it is difficult to deal with conflicting and incommensurable aspects such as environmental, economic and social dimensions. In this paper we have used a Multi-Criteria Decision Aid (MCDA) methodology to cope with these difficulties. MCDA methodology offers the opportunity to avoid monetary valuation of the different dimensions of the SD. These dimensions are not substitutable for one another and all have a role to play. There is an abundance of possible aggregation procedures in MCDA methodology. In this paper we have proposed an innovative method to choose a suitable aggregation procedure for SD problems. Real life case studies of the implementation of an outranking approach (i.e., ELECTRE) and of a mono-criterion synthesis approach (i.e., MAUT approaches based on the Choquet integral) were done to respectively rank 22 SD strategic actions within an expertise Institute and rank 20 practical operational actions to control energy consumption of the Institute's buildings.
Uploads
Papers by Lisa Serir