Fuzzy modeling of dependability optimization for supporting the production-quality strategies - case study in technical field

7th Modern Technologies in Industrial Engineering

工业技术（总论）

Vilcu, A.^1 ; Verzea, I.^1 ; Pislaru, M.^1 ; Herghiligiu, I.^1

Gheorghe Asachi Technical University of Iasi, Department of Engineering and Management, D. Mangeron 63, Iasi, Romania^1

关键词: Artificial intelligence methods;    Artificial intelligence techniques;    Comparison of models;    Complexity of algorithm;    Engineering design process;    Mathematic algorithms;    Quantitative comparison;    Technical equipments;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/591/1/012072/pdf
DOI  :  10.1088/1757-899X/591/1/012072

学科分类：工业工程学

来源: IOP

【 摘 要 】

this work treats dependability from a functional (systemic) standpoint, which implies meeting a function required within an input/output system. This approach is demanded and necessary within technical equipment design stages, and is an integrant part of systems integrity design methodology. Designing the system integrity includes design criteria for reliability, availability, maintainability and safety of any system and equipment. The combination of these four concepts leads to the necessity of a comparative and integrative methodology that should ensure a good systems design, with required integrity values that can be computed easily, analyzed the most complete possible, and with the possibility to be modified accordingly. During the recent years, artificial intelligence techniques have been developed for dependability, that make use of, statistic methods (necessary for the realization of an operation history), deterministic mathematic algorithms (to determine the exact solutions when these can be determined and are required and the time resource is not critical), heuristic artificial intelligence methods (that can provide good quality solutions when time is a critical resource)-neuronal networks, genetic algorithms (to determine predicted or extreme values for system variables), as well as fuzzy methods for systems in which the system variables change within certain values intervals. All these techniques offer a global image of the artificial intelligence modelling (AIM) in designing the reliability, availability, maintainability and safety, in order to offer a continuous design feedback mean during the entire engineering design process. In this work, we shall develop a fuzzy model for a technical process and will compare our results with the results supplied by mathematical, neuronal and genetic models applied to the same system. The qualitative comparison of models will be completed with their quantitative comparison, by analyzing the complexity of algorithms and execution time. The utility of this work consists in the implementation of an inter-disciplinary tool easy to apply to technical systems with measurable dependability variables that can provide during the flow sheet, the values requested from technical and economic standpoint.