Detecting Diamond Breakouts of Diamond Impregnated Tools for Core Drilling of Concrete by Force Measurements

Christine H. Müller (), Hendrik Dohme (), Dennis Malcherczyk (), Dirk Biermann () and Wolfgang Tillmann ()
Additional contact information
Christine H. Müller: TU University Dortmund, Department of Statistics
Hendrik Dohme: TU University Dortmund, Department of Statistics
Dennis Malcherczyk: TU University Dortmund, Department of Statistics
Dirk Biermann: TU University Dortmund, Institute of Machining Technology
Wolfgang Tillmann: TU University Dortmund, Institute of Materials Engineering

A chapter in Artificial Intelligence, Big Data and Data Science in Statistics, 2022, pp 283-300 from Springer

Abstract: Abstract Diamond impregnated tools for core drilling consist of segments in which synthetical diamonds are bounded in a metal matrix. The wear of these tools depends on the time points when active diamonds breakout and new diamonds from deeper layers of the metal matrix become active. Up to now, these time points were measured only by visual inspection at given inspection time points, a measurement which is very error-prone and labor-intensive. Hence the aim is to use the automatic force measurements during the drilling process for detecting the breakouts of the diamonds. These force measurements consist of three time series observed over about 75 min, each minute with over 300,000 measurements. At first, we present here an approach of an analysis of these time series in three steps: identification of the time periods of active drilling, identification of the rotation periods, and determination of differences between successive rotations. Based on the detected rotation periods, 147 features for classification of minutes with and without diamond breakout are created. Some of these features are based on the differences between successive rotations and some on p-values for testing the independence of the detected rotation lengths. After a feature selection step, random forest and logistic regression are applied. This leads at least for one of two considered series of experiments to a classification error which is smaller than the trivial classification error.

Keywords: Classification; LASSO; Logistic regression; Random forest; Dynamic time warping (search for similar items in EconPapers)
Date: 2022
References: Add references at CitEc
Citations:

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:sprchp:978-3-031-07155-3_12

Ordering information: This item can be ordered from
http://www.springer.com/9783031071553

DOI: 10.1007/978-3-031-07155-3_12

Access Statistics for this chapter

More chapters in Springer Books from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().