عنوان مقاله
پشتیبانی از تصمیم برمبنای NPV در طراحی چند هدفه با استفاده از الگوریتم های تکاملی
فهرست مطالب
مقدمه
مسائل پاکسازی داده
شیوه های پاکسازی داده
حمایت ابزاری
نتیجه گیری
بخشی از مقاله
شیوه برتری بر مبنای ارزش
برای طراحی بهینه
شیوه مطرح شده در اینجا، رسیدگی به طراحی بهینه چند هدفه با استفاده از فرمولاسیون مصالحه قیاسی براساس شاخص کل ارزش کل عمر می باشد. این معیار شامل ارزش اقتصادی انتگرال وابسته به محصول بوده و به همین خاطر به عنوان یک شاخص کافی برای برتری کل طرح های انتخابی شناخته می شود. برای این کار، عناصر غیر بازگشتی ( سرمایه گذاریها) و بازگشتی (هزینه های مختلف در طول عملیات) بایستی ذکر شوند
کلمات کلیدی:
NPV-based decision support in multi-objective design using evolutionary algorithms Damir Vucˇina , Zeljan Lozina, Frane Vlak ˇ FESB, Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture, R. Boskovica bb., University of Split, 21000 Split, Croatia ˇ article info Article history: Received 21 November 2007 Received in revised form 29 October 2008 Accepted 30 September 2009 Available online 30 October 2009 Keywords: Decision support Multi-objective Design optimization Net Present Value Multi-layer sandwich plates abstract Optimum design problems are frequently formulated using a single excellence criterion (minimum mass or similar) with evolutionary algorithms engaged as decision-support tools. Alternatively, multiobjective formulations are used with a posteriori decision-making amongst the Pareto candidate solutions. The former typically introduces excessive simplification in the decision space and subjectivity, the latter leads to extensive numerical effort and postpones the compromise decisionmaking. In both cases, engineering excellence metrics such as minimum mass can be misleading in terms of performance of the respective design in the given operational environment. This paper presents an alternative approach to conceptual design where a compound objective function based on the Net Present Value (NPV) and Internal Rate of Return (IRR) aggregate performance metrics is developed. This formulation models the integral value delivered by the candidate designs over their respective life-cycles by applying value-based NPV discounting to all objectives. It can be incorporated as an a priori compromise and consequently viewed as a weighted sum of individual objectives corresponding to their economically faithful representation over the entire operational life-time of the designs. The multi-objective design optimization is consequently expanded from purely en
