Development of a rough set-based decision support system for life cycle impact assessment and interpretation

• Main Author: Aviso, Kathleen B.
• Format: text
• Language: English
• Published: Animo Repository 2006
• Subjects: Decision support systems; Rough sets; Product life cycle > Environmental aspects; Environmental Engineering
• Online Access: https://animorepository.dlsu.edu.ph/etd_masteral/3372
• Institution: De La Salle University

Life cycle assessment (LCA) is a methodological framework for assessing the environmental impacts of products or processes during their entire lifetime. It consists of four phases: (1) goal and scope definition, (2) inventory analysis (LCI), (3) impact assessment (LCIA) and (4) interpretation. LCA involves the simultaneous evaluation of multiple criteria or multiple goals. A systematic way of dealing with this problem is provided by Decision Analysis techniques, particularly through the use of multiple criteria decision analysis (MCDA) methods. MCDA methods include the multi-attribute utility/value theory (MAUT/MAVT), outranking methods and the analytical hierarchy process (AHP). Thus recognizing the benefits, most of the existing LCIA and interpretation methods patterned their frameworks to MCDA methods. However, these require decision makers (DM) to express their preferences into importance weights or parameters, which are necessary for the chosen preference model - a task, which is tedious. Hence, an alternative approach is recommended in this study. The use of rough set methodology has been successfully applied to multiple criteria or multiple attribute problems in engineering, medicine, banking, economics, and financial and market analysis. It is capable of finding patterns in data and dealing with uncertainties and inconsistencies, which may be due to a DMs limited discriminatory power. It only requires previously expressed decisions made by the DM to infer the DMs adapted preference model in terms of decision rules. This study thus presents the development of a decision support system (DSS) utilizing a two-step procedure of Pareto optimality and rough set methodology for impact assessment and interpretation. This alternative methodology has shown comparability in results with AHP and was found to predict accurately the decisions of experts to a degree of 83%. The model, which is founded on the decision rules derived from the assessment of a panel of experts on a set of power generating technologies, encapsulates the environmental concerns considered and the state of knowledge of the experts during the time of the survey. Thus this model can be utilized to rank and evaluate new technologies against four other systems, which are stored in the models database, based on the same arguments utilized for assessing the training data examples.