Από κοινού παρακολούθηση της μέσης τιμής και της μεταβλητότητας μιας διεργασίας με διαγράμματα ελέγχου ροής αθροίσματος
Joint monitoring of the process mean and variability with run sum control charts
View/
Keywords
Διάγραμμα ελέγχου ; Ενιαίο διάγραμμα ελέγχου ; Μέσος και μεταβλητότητα της διεργασίας ; Μέσο μήκος ροής ; Μαρκοβιανή αλυσίδαAbstract
In the present PhD dissertation new single control charts are presented for the joint
monitoring of the parameters of a production process.
Specifically, the focus is on the study and development of single run sum control
charts which are capable of detecting shifts in process parameters by plotting a single
statistic. First, a single run sum control chart is studied for the simultaneous monitoring
of the mean and the standard deviation of the process. The proposed control chart is
suitable for detecting increases or decreases in the mean and/or the standard deviation
of the process. Next, a adaptive single run sum control chart is presented and studied
for the joint monitoring of the mean and/or standard deviation. In this chart the sample
size and/or the sampling interval may vary, in order to further increase the sensitivity
of the control chart to detect shifts in process parameters. Finally, a run sum control
chart is presented to simultaneously monitor the origin and scale parameter of a shifted
exponential process, where the distribution of the critical-to-quality characteristic
follows the shifted exponential distribution. The proposed control charts are easy to use
and able to identify the parameter or parameters that have been shifted.
From the relevant literature review, it is emerged that the charts proposed and
studied in this PhD thesis are original. Moreover, the numerical analysis showed that
they can be reliable joint monitoring procedures, as they have improved performance
in detecting a variety of out-of-control situations. For all the above-mentioned control
charts we present their statistical design and operation. Their performance is calculated
when we know a priori the magnitude of the shifts in process parameters. Also, their
overall performance is calculated, for a wide range of shifts in process parameters.
Finally, the practical implementation of the proposed charts is demonstrated to real
industrial problems.
Department
Σχολή Χρηματοοικονομικής και Στατιστικής. Τμήμα Στατιστικής και Ασφαλιστικής ΕπιστήμηςNumber of pages
287Language
GreekCollections
Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα