A reverse logistics planning problem is modeled and analyzed. The model considers returns of a particular electronic device from customers. Some of the collected products are remanufactured or refurbished. Others are disassembled for their key parts which can be considered as good as new. New products are assembled either using new parts or extracted ones. There are two types dynamic demands: demands for remanufactured/refurbished products and demands for new products. Demand of remanufactured/refurbished products can be satisfied using new products in case of shortage. This is a one way downward substitution. The objective is to minimize total costs while satisfying all demands. This problem is formulated as a MILP. The numerical results show that: i) it is hard for a solver to find optimal solutions for the problem in reasonable computational times for several instances with relatively small time horizons and ii) substitution is justified for a certain range of cost and demand parameters.

It is always difficult to identify the most recent works that have been published, especially those published in recent years, due to delays in putting publications online, citations indexe, etc. Scientometry offers to researchers various concepts, models and techniques that can be applied to knowledge management (KM) in order to explore its foundations, its state, its intellectual core, and its potential future development. To this end, we have developed a scientometric KM framework to calculate the scientometric indexes related to a query introduced in the Scopus database, to facilitate research and monitoring of productivity and collaboration between the authors of KM in particular and also the dissemination of knowledge. The works between 2014 and 2019 are taken, the industry of services was omitted. It might help the decision makers and researchers to optimize their time and efforts. We used Unified Modeling Language (UML) to translate the development ideas of the scientometric framework structure into diagrams, and Delphi 7 to calculate the indexes and ensure other operations of research (about: articles, their authors, conferences, etc). This framework is only valid for Excel files extracted from Scopus or similar format. Finally, the relation between KM and industry 4.0 was established on found articles in Scopus.

In this paper, a direct vector control using fuzzy logic controller with adaptive gain for a doubly fed induction generator (DFIG) based wind energy conversion system (WECS) is presented. The performance of fuzzy controllers is characterized by unsatisfactory performance: (wide overshoot, excessive oscillations and sensitivity to parametric variations). We propose a robust method, where the control gain will be continually adapted with the use of a set of fuzzy rules; we only consider the gain adaptation of the command. I mean the value of the gain will be determined by a rule base defined by the error and the variation of the error. Finally, the control of the active and reactive powers using a fuzzy logic controller with adaptive gain is simulated using software Matlab/Simulink, studies on a 1.5 MW DFIG wind generation system compared with the conventional fuzzy logic controller. Performance and robustness results obtained are presented and analyzed. KEY WORDS Wind energy conversion system ; Vector control ; Fuzzy logic controller ; Adaptive fuzzy logic controller.

The increasing complexity of most industrial processes always tends to create problems in monitoring and supervision systems. Detection and early fault diagnosis are the best way to manage and solve these problems. Artificial neural networks (ANNs), by their ability to learn and store a large volume of information, are tools particularly suitable for diagnostic support systems. Effectiveness of ANNs for fault diagnosis in milk pasteurization process is presented in this paper. The initial data base used for fault diagnosis is constructed using data extracted from FMEA (Failure Modes and Effects Analysis) tables of milk pasteurization process. Indeed, this analysis makes it possible to establish the links of cause and effect between the faulty components and the observed symptoms. Three models of ANNs, namely Feed-Forward Back Propagation (FFBP), Radial Basis Function based Neural Network (RBNN), and Generalized Regression Neural Networks (GRNN) are developed and compared. The determination coefficient (R2 ), Root Mean Square Error (RMSE), and Mean Absolute Error (MAE) statistics were used as evaluation criteria of all the models. The comparison results indicate that the performances of GRNN model are better than the FFBP and RBNN models. The same neuronal models can be extended to any technical system by considering appropriate parameters and defects.