A Fuzzy - Based Methodology for Aggregative Waste Minimization in the Wine Industry.
[Book]
Ndeke Musee
INTECH Open Access Publisher
2010
In summary, the proposed knowledge-based decision support system provides a systematic approach for evaluating and diagnosing the unstructured WM problem in the wine industry by way of processing the user inputs (both qualitative and quantitative) to compute a given winery performance in terms of effluent quantity, product and by-products recovery, chemical usage, or effluent quantity. The system results are in a format that can be easily read, understood, or altered by the user. This is because the final system outputs are expressed in the form of performance indexes (range [0, 1]), and therefore, the proposed decision model offers a transparent and robust tool for assessing the performance of a given winery with respect to WM. Secondly, the system incorporated data, information and knowledge sourced from experts that can aid in facilitating efficient decision-making regarding WM in the wine industry. Thirdly, as the wine industry has dearth of statistical data regarding WM unlike the chemical industry, fuzzy logic and qualitative reasoning soft computing approaches were applied to aid in evaluating WM in this industry. Presently, there is no evaluation-framework tool that can assess WM performance of a given winery through integration of both qualitative and quantitative data. Thus, the integrated methodology proposed in this paper serves a suitable tool to achieve this objective as well as to aid in automating WM analysis in the wine industry. Consequently, based on the integrated framework presented, the WM can be evaluated and ranked at different levels of aggregation clearly identifying areas that may need improvement to optimise resources utilization and reduce operational costs. And finally, the system has the merit of reducing the time, effort, and resources required in undertaking extensive WM in the wine industry. The suitability of the approach has been demonstrated through two worked case studies, each with several different functional scenarios.