Suatu perbandingan antara carta-carta kawalan berstatistik dengan carta kawalan berdasarkan rangkaian neural

Abstract

Control chart is one of the most powerful quality tools in statistical process control and is widely used in the manufacturing process. As the demand of the quality control increases, traditional control chart is no longer sufficient to detect the sudden change in a process. Thus, run rules are built-in into the Shewhart X control chart. This improvement, and also EWMA charts are introduced to overcome the limitation to its sensitivity. Neural network based control chart is a new approach and its performance is compared with the statistical control charts. The same test data from the standard normal random variable is generated using SAS. This data is assumed as the in-control data for the three types of control charts. The criteria to compare the performance of both types of control charts is the average run length (ARL). From the results obtained, neural network has a better ARL than the statistical control charts which includes the run rules of Shewhart X control chart and the EWMA chart when detecting small and large shifts in the process mean