Controlling processes with reference to costs, item price and process evolution

This thesis presents some recent work of the author in developing the analysis of a number of process control models that take into account statistical, economic and other practical issues. Special attention is paid to the problem of optimum selection the initial process mean setting, with particular reference to filling/canning processes. As there are many different situations that involve different cost parameters, this leads to the consideration of various models each with their own particular solution. The effects of change of the process variance on the optimal solution as well as on the expected profit are discussed. Implications to 'Weights and Measures' requirements of following this optimality path are provided, with particular reference to loss in expected profit per item. Chapter 1 provides a brief introduction and is followed by a literature review in Chapter 2. Chapter 3 deals with the issue of selecting the optimum process mean by presenting a simple model and emphasising the dependencies between the process parameters. Chapter 4 further investigates the problem presented in Chapter 3 and presents several models for which the selection of the most profitable process setting is considered, concentrating on a canning problem. Various industrial filling processes are described and some of the issues considered include: waste, overfill, top-up, and the penalty costs for items that initially fail to meet specifications. Chapter 5 discusses Weights and Measures requirements in connection with a canning process. Both Australian requirements and OIML International recommendations are discussed. The Australian requirements are also compared with the requirements of the European Economic Community as well as the United States. Chapter 6 illustrates the potential use of the models developed in chapter 4 by giving an industrial example and again discussing the implications to Weights and Measures requirements. In Chapter 7 the problem of an optimal selection of the initial process mean is examined for a process with a linear shift. Special focus is on the economic benefits obtained from reducing process standard deviation and the rate of change of the mean. Conclusions and some suggestions for future work are provided in chapter 8. Parts of chapter 4, 5 and 6 form the contents of a paper, 'Mean Selection for Various Types of Filling Process with Implications to 'Weights and Measures' Requirements', undergoing revision for publication in the Journal of Quality Technology