Multivariate control charts and applications

The control charts which are used as the systematic approach of the quality control methods are created by the transfer of the measurement results taken from the identified quality characteristics or the statistics generated from them into the charts. In the real life problems, the cases where there are two or more correlated quality characteristics are unavoidable. In the resolution of the multivariate control problems in the literature, the control chart methods of Hotelling T², MCUSUM and MEWMA become prominent. Among these methods, the control chart method of Hotelling T² is the most preferred method due to its ease of application. While information about whether the multivariate process is under control is provided by means of this control chart method, the information on from which variable or variables this out-of-control status arises cannot be provided. Certain decomposition methods which are complementary for these control chart methods and which are used for the variability to be able to be determined have been developed. Among these methods, MYT decomposition method is one of the methods which are most commonly used and which ensure that the most accurate results are obtained. In this thesis study, the examination of the top rack of a dishwashing machine which was produced at a dishwashing machine factory and into which the dishes are placed was made due to its customer complaints. In the use of the top rack, 4 quality characteristics which lead to inferiority were identified by evaluating the customer feedbacks. The daily measurements were taken for these quality characteristics and the assessment of the out-of-control status was made by applying the control chart method of Hotelling T² from the multivariate control methods. In order to be able to determine from which quality characteristics of the out-of-control status, i.e. the variability, arises, however, MYT decomposition method was used.