Monitoring strategies for chatter detection in a drilling process

Abstract

The thesis is divided into two main parts. The first part is devoted to the development of a new distribution-free control chart for multivariate data, the study of its properties and comparison of its performance against alternative procedures. In Chapter 2, the principles used to construct the new control chart are given and its statistical design is proposed. Simulations are conducted in Chapters 3 and 4 in order to assess its applicability and to compare its performance against alternative procedures, respectively. The second part of the thesis is devoted to the introduction and modelling of the drilling process and experimental investigation of the performance of the proposed monitoring strategies. In Chapter 5, the BTA deep-hole drilling process is introduced. Previous work to understand and to predict the temporal development of the process is reviewed in Chapter 6. Based on it, a modelling approach of the sampled data is proposed in order to setup the monitoring procedures. In Chapter 7, nonlinear time series modelling is used to describe the time varying dynamics of the process. It provides the basis of nonlinear time series based control charts. In Chapter 8, the empirical performance of the proposed monitoring strategies is studied using several data sets. Finally, some thoughts and discussion for possible future research issues are given in Chapter 9.