Computational intelligence for predictive condition monitoring and approaches for online analysis

Abstract

‎The process of high speed milling (HSM) is regarded as one of the most sophisticated and complicated manufacturing operations‎. ‎In the past four decades‎, ‎many investigations have been conducted on this process aiming to better understand its nature and‎ ‎improve the surface quality of the products as well as extending tool life‎. ‎To achieve these goals‎, ‎it is necessary to form a‎ ‎general descriptive reference model of the milling process using experimental data‎, ‎thermo-mechanical analysis‎, ‎statistical or‎ ‎artificial-intelligent (AI) models‎. ‎Besides‎, ‎increasing demands for more efficient milling processes‎, ‎qualified surface finishing‎, and modeling techniques have propelled the development of more effective modeling methods and approaches‎. ‎In the first part this dissertation‎, ‎an‎ ‎extensive literature survey of the state-of-the-art modeling techniques of milling processes is carried out‎, ‎more specifically‎ ‎of recent advances and applications of AI-based modeling techniques‎. ‎The comparative study of the available methods as well as the‎ ‎suitability of each method for corresponding types of experiments is also presented‎. ‎In addition‎, ‎the weaknesses of each method as‎ ‎well as open research challenges are presented‎. ‎Therefore‎, ‎a comprehensive comparison of recent developments in the field will‎ ‎be a guideline for choosing the most suitable modeling technique for this process regarding its goals‎, ‎conditions‎, ‎and specifications‎. ‎After comprehensive study of the available methods for modeling HSM processes‎, ‎to build up a proper condition monitoring system‎, ‎sensor signals are to be utilized to form a reference model which non-intrusively reflects the performance of the system‎. ‎Therefore‎, ‎a desired reference model has to apply more efficient feature extraction and artificial intelligence (AI) techniques to be more repeatable and generalizable‎. ‎Since milling signals are complex‎, ‎a time-frequency analysis method‎, ‎namely wavelet‎, ‎is applied for feature extraction‎. ‎Considering the high dimension of the wavelet features‎, ‎clustering methods are used for dimension reduction and also as an interpretation layer between the signal feature extraction subsystem and artificial intelligence blocks‎. ‎This research illustrates the performance of artificial intelligence based techniques for modeling of high speed end milling experimental data‎. Studied and developed methods are applied on wavelet features of force and vibration signals to illustrate the repeatability and accuracy of their results‎. ‎It is shown that the proposed structure as well as the developed artificial intelligent method can present the status of the process and can be applied for fault diagnosis and TCM purposes‎. ‎It is also discussed that how application of available data mining methods with a proper structure may improve the performance of existing reference models towards more efficient utilization of available experimental data and easily generalizable reference models‎.