Please use this identifier to cite or link to this item:
|Title:||NEURAL NETWORK APPROACH FOR ANALYSIS OF CONTROL CHART PATTERNS|
|Keywords:||MECHANICAL INDUSTRIAL ENGINEERING;NEURAL NETWORK APPROACH;CONTROL CHART PATTERNS;SHEWHART CONTROL CHART|
|Abstract:||Shewhart control chart is an essential tool in statistical quality control. The power of Shewhart technique lies in its ability to separate out assignable causes of quality variations . Pattern recognition task is an important aspect of interpretation of Shewchart control charts. Previous researches in control chart were primarily concerned with the detection of shifts in the process mean. There are many other patterns which may exist in process data indicating out of control situation. When these patterns occur , analysis of the control charts become a pattern recognition problem. Over the years numerous supplementary rules, have been proposed to analyze the control charts. These rules were developed to assist operators for detection of unnatural patterns. The interpretation of process data still remains difficult because these involve pattern recognition aspects This study deals with pattern recognition problem . A method incorporating Error Back Propogation —ANN is proposed. to make possible the analysis of process data in real time with little or no human intervention|
|Research Supervisor/ Guide:||Arora, Navneen|
|Appears in Collections:||MASTERS' THESES (MIED)|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.