دانلود مقاله ISI انگلیسی شماره 15237
ترجمه فارسی عنوان مقاله

نظارت پیشگویانه غیرمستقیم در سیستم های تولید انعطاف پذیر

عنوان انگلیسی
Indirect predictive monitoring in flexible manufacturing systems
کد مقاله سال انتشار تعداد صفحات مقاله انگلیسی
15237 2000 18 صفحه PDF
منبع

Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)

Journal : Robotics and Computer-Integrated Manufacturing, Volume 16, Issue 5, October 2000, Pages 321–338

ترجمه کلمات کلیدی
- سیستم های تولید انعطاف پذیر - تعمیر و نگهداری پیشگویانه - نظارت پیشگویانه غیرمستقیم - جریان تولید - شکست - سرعت - شناسایی - تشخیص رانش -
کلمات کلیدی انگلیسی
Flexible manufacturing systems, Predictive maintenance,Indirect predictive monitoring,Production flow,Failures,Drift rate,Detection,Diagnosis
پیش نمایش مقاله
پیش نمایش مقاله  نظارت پیشگویانه غیرمستقیم در سیستم های تولید انعطاف پذیر

چکیده انگلیسی

In this paper, a new method of monitoring of failures in flexible manufacturing systems (FMS) is developed. The main objective is to manage progressive failures in order to avoid breakdown state for FMS. To achieve these requirements an indirect predictive monitoring approach is proposed. The main ideas of this approach are first to follow the state evolution of the resources indirectly by the perturbations on the manufactured products. The parameter used is the production flow for which a drift rate indicator is defined for each of the family. Detection and diagnostic methods based on the on-line exploitation of the drift rates are presented in this paper. Another idea developed is the necessity to integrate different approaches of monitoring to obtain a global management of failures. For this purpose a functional framework of integrated monitoring in the control command of an FMS is proposed.

مقدمه انگلیسی

The actual evolution of flexible manufacturing systems tends towards just in time by production control strategies enabling the elimination of excesses, rejects and variations. However, the occurrence of failures during the exploitation stage can deeply modify the FMS performances [1] or its availability. Since 10 years the L.A.I.L is interested in developing methods to manage the failures of FMS processes. The work developed in the field of monitoring of FMS is a part of a global project to develop methodologies to assist the design of FMS control/command: it is CASPAIM project [2]. Traditionally, two types of failures are considered in dependability: cataleptic and progressive failures. Cataleptic failures are failures that cannot be detected by process inspections achieved by maintenance operators; they are sudden and complete. According to their impact, they induce immediately a breakdown state of the faulty components or system. At the opposite, progressive failures characterise degradations occurring in a process. Consequently, they are partial. Their evolution can be followed and recovery actions [3], [4] and [5] can be performed before the breakdown state is reached.

نتیجه گیری انگلیسی

In this paper, a new approach called indirect predictive monitoring has been introduced. It enables to survey FMS; the main idea is to analyse the information on resources’ states that are transported by manufactured parts themselves. In this paper, only the quantitative approach has been focused by the analyses the process flows. It is important to understand that this idea can also be applied considering the quality of the manufactured parts. Concerning this study, the analysis of the process flows is performed with the concept of drift rate. It permits to compare flows some different point of views: regardless of planned scheduling, regardless of the difference between the number of parts entering in a sub-set of the FMS and exiting from this sub-set. In this study, the second point of view has been considered after the introduction of two types of reasoning: reasoning at N constant that is useful to survey the cycle time of individual resources and reasoning at T constant which is useful to survey the behaviour of a set of resources for which the cycle time is not easily available. The valuation part has shown that the drift rate enables to detect failures. However, induction phenomena make the diagnosis more complex. To avoid physical induction, the diagnosis must be performed quickly after the detection of the failures. As soon as the faulty resource has been identified, the pilot must adjust the frequencies of parts entries in order to avoid the saturation of buffers. The value of elementary period used by the acquisition process must also be adjusted to avoid operating sequence induction.