| As manufacturing quality has become a decisive factor in global market competition, statistical process control (SPC) and engineering process control (EPC) are widely used in manufacturing industry. My research is motivated by the characteristics of modern manufacturing environments where the life cycle of products has decreased rapidly.;Excluding the Introduction part and Literature Review part, this thesis consists of three parts. The first part is Chapter three which is a topic on SPC. I develop two definitions for the unbiasedness of a control chart and indicate that one of self-starting control charts, Shewhart Q -chart, would be biased if the process mean shifts very early and the variance of the process is unknown. In order to alleviate the bias problem of the Shewhart Q-charts, two improved control charts are proposed and investigated through mathematical proof and Monte Carlo simulation.;The second part includes Chapter four and Chapter five which are topics on EPC. Chapter four develops a new controller called general harmonic rule (GHR) controller. The GHR controller is designed based on the assumption that the initial bias is an unknown and non-random value. It is shown that the GHR controller is more robust than the EWMA controller when the process parameters have uncertainty. In Chapter five, I assume the initial bias is a random variable and build a state-space representation for the case that the disturbance is assumed to be IMA(1,1) or ARMA(1,1) process. An optimal control algorithm is presented in Chapter five and the proposed formulation is Bayesian and based on Kalman filter.;The last part is Chapter six, which is a topic on an integrated EPC and SPC approach. I assume the relationship between input and output of the process to be controlled is unknown. In the EPC procedure, each output is assigned an influence score based on the deviation of the output from the target. The rule to assign the influence score is according to an influence function. In the SPC procedure, an EWMA control chart is used to monitor the process outputs. The performance of the integrated EPC and SPC control scheme is presented through Monte Carlo simulation when the family of power influence functions is used. Currently, this integrated EPC and SPC approach is only a rough framework. Future research is mentioned at the end of Chapter six. |
