پديد آورنده :
مقدم، غزاله
عنوان :
توسعه روش هايي براي مراحل اول و دوم پايش پروفايل خطي ساده فازي
گرايش تحصيلي :
مهندسي صنايع
محل تحصيل :
اصفهان: دانشگاه صنعتي اصفهان، دانشكده مهندسي صنايع و سيستم ها
صفحه شمار :
چهارده، 179ص.: جدول، نمودار
استاد راهنما :
غلامعلي رئيسي اردلي
استاد مشاور :
وحيد اميرزاده
توصيفگر ها :
آماره فازي اي دبليو ام آ , آماره فازي تي دو هتلينگ , رگرسيون فازي , آزمون فرض فازي , تكنيك نقطه تغيير فازي
استاد داور :
علي همداني، ماشاالله ماشين چي، عباس پرچمي
تاريخ ورود اطلاعات :
1395/02/07
دانشكده :
مهندسي صنايع و سيستم ها
كد ايرانداك :
ID874 دكتري
چكيده انگليسي :
Developing methods for phase I and phase II fuzzy simple linear profile monitoring Ghazale Moghadam g moghadam@in iut ac ir Date of defense Department of industrial engineering Isfahan University of Technology Isfahan Iran Degree Ph D Language Farsi Supervisor Gholam Ali Raissi Ardali raissi@cc iut ac ir Abstract In most of the applications of quality control the quality of a process or a product isdescribed by a relationship between the response variable and one or several independentvariables Such relationship is called a profile Profile monitoring consists of two phases The goal of phase I is to model process s performance and estimate model s parameters in astable situation The goal of the phase II is to promptly identify changes in the profile andthe causes of such out of control states In most real world issues vagueness imprecisionand uncertainty in information are unavoidable Hence in this research we investigate andpropose methods for process monitoring in phases I and II of processes with linear profilesin controlled situations with vague and fuzzy observed response variables The proposedmethod for phase I of monitoring simple linear profiles with fuzzy observations is based onfuzzy change point technique The efficiency of the method is evaluated by means ofsimulation and based on the probability of out of control signal criterion Results show thatnot only is this method efficient but it has also the advantage over competing methods thatit provides the analyst with an estimate of the point of divergence of the process fromcontrol This feature enhances identification and deletion of out of control points to get aset of in control data Two methods are proposed for the phase II of profile monitoring with fuzzyobservations In these methods initially the profile parameters are determined by means offuzzy regression technique Afterwards fuzzy EWMA and fuzzy Hotelling s T statisticsand fuzzy hypothesis testing are employed To this end fuzzy EWMA and fuzzy Hotelling s T statistics are developed using extension principle The efficiency of methods aretested by means of simulation and based on the average run length The results show thatthe proposed method based on fuzzy EWMA statistic and fuzzy Hotelling s T statistic arerespectively efficient for identification of small changes and for identification of mediumand significant changes Considering the sensitivity of the process and the goals of theanalysis it is possible to use either one of the methods or a combination of the two tomonitor profiles of processes that their quality characteristics are vague imprecise anduncertain Keywords fuzzy simple linear profiles fuzzy EWMA statistic fuzzy Hotelling s T statistic fuzzyregression fuzzy hypothesis testing fuzzy change point technique
استاد راهنما :
غلامعلي رئيسي اردلي
استاد مشاور :
وحيد اميرزاده
استاد داور :
علي همداني، ماشاالله ماشين چي، عباس پرچمي
