General Information
1. Summary
The project considers the Change Detection and Diagnosis problem (CDD) in vibrational processes using advanced measuring and analysis techniques model-based. The vibrational processes are characterized by vibrational phenomena, which include mainly, as effect, mechanical vibration signals, resulting in normal or abnormal operating.
It is an important trend concerning the replacement of the systematic procedure of maintenance of machinery and equipment's by conditional maintenance strategies, based on continuously or selective monitoring of the process, with the scope of detection of abnormal behaviour and to avoid catastrophic events of economic or ecological nature. In this context, early time detection of abnormal behaviour of systems seems to be a necessary solution, possible and efficient, in rapport with a right working mode description, without artificial excitation, change of the working regime or breaks.
2. General Objectives
The general objective of the project is to build an experimental model for CDD with application in vibrational process monitoring, using advanced measuring and analysis techniques model-based.
3. Specific Objectives
The specific objectives are: (O1) Development, implementation and validation of new methods, techniques and algorithms for CDD; (O2) Optimization of classical algorithms for CDD; (O3) CDD information fusion coming from and in time of process monitoring; (O4) Development of a CDD software library, which will implement both classical (known) methods and optimized and new developed ones during the project running; (O5) Development and testing of an experimental model for CDD, with commercial features, hardware and software, which will use the results obtained during the project, under all aspects: theoretic, algorithmic and methodologic. The monitoring of the vibrational processes will consider, in project, another two waves, which are naturally generated and accompanying, partial or complete, continuously or discontinuously, the mechanical vibrations: (1) acoustic waves, inside the audio spectrum; (2) ultrasound waves, with frequencies up to 100 kHz. The project looks on information analysis and processing generated by the set of three presented sources, and also to information fusion, for the best decision. The approach will allow the improvement of the process monitoring, more efficient and matched to the considered scope, in rapport with non-fusion, and will be a novelty on national level and one of the few approaches in international area. The project is sustained by the following consortium: "Dunarea de Jos" University of Galati (Coordinator); National Institute of Research and Development in Informatics, Bucharest (Partner 1); National Institute of Research and Development in Mechatronics and Measurement Technique, Bucharest (Partner 2); Teamnet Engineering SRL of Galati (Partner 3). The project will build two products, both new, original and international competitive, which will provide solutions to CDD problems of vibrational processes. P1: A program library, as Toolbox of Matlab, which will implement the best algorithms for CDD, using both classical and advanced techniques, as those based on multiresolution analysis, information fusion and soft computing. The product will build a reference for CDD problem and will allow the performance evaluation of new algorithms to the old ones. By using real data from vibrational processes, CDD benchmarks will be proposed; P2: an experimental model, having a CDD software application as basis, to be used in monitoring of some pilot processes, in laboratory, and of a complex industrial process, a thin strip mill, at S.C. ArcelorMittal S.A. Galati. The physical model will be the basis to launch full commercial products for various processes and markets. The presented results, as well as the support systems resulted (e.g. measuring and testing systems) will rise the procedural algorithmic level for the improvement of industrial process monitoring, enhancing their operational safety.
4. Work Plan
Initial (01.07.2015)
Work Plan (01-07-2014/31-06-2016)