2 results
Search Results
Now showing 1 - 2 of 2
Conference Object Citation - Scopus: 2Comparison of Gross Calorific Value Estimation of Turkish Coals Using Regression and Neural Networks Techniques(2012) Ozbayoglu,A.M.; Ozbayoglu,M.E.; Ozbayoglu,G.Gross calorific value (GCV) of coals was estimated using artificial neural networks, linear and non-linear regression techniques. Proximate and ultimate analysis results were collected for 187 different coal samples. Different input data sets were compared, such as both proximate and ultimate analysis data, and only proximate analysis data and only ultimate analysis data. It was observed that the best results were obtained when both proximate analysis and ultimate analysis results were used for estimating the gross calorific value. When the performance of artificial neural networks and regression analysis techniques were compared, it was observed that both artificial neural networks and regression techniques were promisingly accurate in estimating gross calorific values. In general, most of the models estimated the gross calorific value within ±3% of the expected value.Conference Object Citation - Scopus: 2Evaluation of Partner Companies Based on Fuzzy Inference System for Establishing Virtual Enterprise Consortium(Springer Verlag, 2015) Nikghadam,S.; LotfiSadigh,B.; Ozbayoglu,A.M.; Unver,H.O.; Kilic,S.E.Virtual Enterprise (VE) is one of the growing trends in agile manufacturing concepts. Under this platform companies with different skills and core competences are cooperate with each other in order to accomplish a manufacturing goal. Success of VE, as a consortium, highly depends on the success of its partners. So it is very important to choose the most appropriate companies to enroll in VE. In this study a Fuzzy Inference System (FIS) based approach is developed to evaluate and select the potential enterprises. The evaluation is conducted based on four main criteria; unit price, delivery time, quality and past performance. These criteria are considered as inputs of FIS and specific membership functions are designed for each. By applying fuzzy rules the output of the model, partnership chance, is calculated. In the end, the trustworthy of the model is tested and verified by comparing it with fuzzy-TOPSIS technique providing a sample. © Springer International Publishing Switzerland 2015.
