Statistical methods for condition monitoring systems
The use of sensor networks for monitoring technical systems in real time, and consequently providing information about the condition of a system and providing decision support for maintenance, inspection and repair strategies, is common in various industries and transport systems. In the maritime industries, there has also lately been much interest in condition monitoring systems as a means to ensure the safety of shipping. The layout of the sensor network and the collection and storage of sensor data are important issues to consider in a condition monitoring system. However, one equally important task is the reliable and timely analysis of the available data, the analytics of the system, in order to extract useful information from the sensor signals. The types of analytics that are most relevant for a condition monitoring system are normally referred to as diagnostics and prognostics. Diagnostics typically refers to assessing the current state of the system, whereas prognostics looks ahead and predicts the future development of the system, for example by predicting the remaining useful life of the system. Two key approaches to analytics in a condition monitoring system are model-driven approaches and data-driven approaches. Model-based analytics approaches are based on the physics of the system and a model description from first principles. Data-driven approaches, on the other hand, utilise the information in a knowledge base or available data and are essentially statistical methods. Hybrid approaches denote a combination of model-driven and data-driven approaches. The focus of this paper is on data-driven methods, specifically statistical analytical methods relevant to condition monitoring of ship machinery systems. This paper presents an introduction and high-level review of a number of statistical methods and approaches that are relevant for diagnostics and prognostics of ship machinery systems based on sensor data from a condition monitoring system. In particular, different statistical methods for classification, which is a major task in diagnostics, are presented.
No Reference information available - sign in for access.
No Citation information available - sign in for access.
No Supplementary Data.
No Article Media
Document Type: Research Article
Publication date: February 1, 2018
More about this publication?
- IJCM is a scientific-technical journal containing high-quality innovative in-depth peer-reviewed papers on all the condition monitoring disciplines, including: acoustic emission methods, electric motor insulation and signature analysis, flow rate monitoring, infrared thermography, lubrication management, optical monitoring, pressure monitoring, temperature monitoring, vibration analysis and also on damage and failure analysis, modelling for condition monitoring, prognostics, sensors and actuators.
- Information for Authors
- Submit a Paper
- Subscribe to this Title
- Membership Information
- Terms & Conditions
- Ingenta Connect is not responsible for the content or availability of external websites