Leila Hayet Mouss

Modeling and evaluation of single machine flexibility using fuzzy entropy and genetic algorithm based approach

Flexibility has long been recognized as a manufacturing capability that has the potential to impact mainly the competitive position of an organization. The entropy approach, which was extended from information theory, fell in handling problems with incomplete and uncertain data, because it depicts only the stochastic aspects included with measured observations. In order to get a global view, this work proposes a new approach based on fuzzy entropy concept. The development of the fuzzy model results in a set of nonlinear constrained problems to be solved using a metaheuristics method. The applicability of our approach is illustrated through a flexible manufacturing cell. By adopting such framework, both dimensions of uncertainty in system modeling, expressed by stochastic variability and imprecision, can be taken into consideration.

Fault diagnosis for a milk pasteurisation plant with missing data

This paper addresses the problem of fault diagnosis from observed data containing missing values amongst the inputs. In order to provide good classification accuracy for the decision function, a novel approach based on support vector machine and extreme learning machine is developed. SVM mixture model is used to model the data distribution, which is adapted to handle missing values, while extreme learning machine enables to devise a multiple imputation strategy for final estimation. In order to prove the efficiency of our proposed method, we have developed the classifier using the condition monitoring observations from milk pasteurisation data. The experiments show that the proposed algorithm gives improved results compared to recent methods, essentially if the number of missing data is significant. The results show that our approach can perfectly detect dysfunction, identify the fault, and is strong in unsupervised process monitoring.

A novel approach for multivariate process monitoring using several intelligences

This paper presents a multi-agent system for multivariate process monitoring. The proposed multi-agent system combines several intelligences which are: multivariate control charts, neural networks, Bayesian networks, and expert systems. This system aims to realise a complete control of complex industrial process. In order to demonstrate the efficiency of the proposed multi-agent system, it has been applied and evaluated in the monitoring of the complex process Tennessee Eastman process (TEP).

Formal approach based on petri nets using agent paradigm for m-maintenance

The long use of a system in a manufacturing environment causes its degradation, thus the maintenance activity is required in this environment to keep and to improve the efficiency of the system. The new development in networking technologies enhances maintenance strategies and gives birth to remote maintenance (tele-maintenance, e-maintenance, m-maintenance). This maintenance makes information available anywhere/anytime and provides maintenance-personnel with the necessary information at the suitable time. This new type of maintenance reduces the maintenance costs and solves the problem of the unavailability of experts. Mobile agent as a rich design concept brings many facilities in the development of m-maintenance, however few works are elaborated in this stage. The objective of this work is both: 1) the proposition of a based mobile multi-agent architecture dedicated for m-maintenance in manufacturing systems; 2) the exploitation of high level petri nets in the specification, simulation and verification phases of the architecture development.

The use of artificial immune system algorithms in monitoring industrial

We describe in this paper an overview of artificial immune system algorithms to solve the classification problem in industrial monitoring. We present artificial immune system algorithms, starting with the negative selection that happens to be a rich source of inspiration. We also, detail the clonal selection algorithm, which is based on the clonal selection theory. Finally, we detail other algorithms based of agent including the immune system and dendritic cell algorithm. In the end, we summarize the differences and similarities of the works discussed and we conclude on the prospects related to the approach of the algorithms of artificial immune systems for industrial monitoring to solve the classification problem.