Back to search

NAERINGSPH-Nærings-phd

Statistical methods for high dimensional sensor based monitoring of ship systems

Alternative title: Statistiske metoder for høydimensjonal sensorbasert monitorering av skip

Awarded: NOK 1.6 mill.

Project Number:

251396

Project Period:

2015 - 2019

Funding received from:

Organisation:

The main aim of this thesis is to develop data-driven and sensor-based models, and to study how these can be implemented and applied in safety critical domains. When the consequence of faulty predictions are low, the path from algorithm and model development to full scale implementation can be relatively short. For such applications, increased accuracy is often enough to justify implementation. For safety critical applications however, trust and robustness are essential, and interpretations and explainability can sometimes be a prerequisite. The thesis is conducted in close collaboration with DNV GL, a global quality assurance and risk management company. It provides classification, technical assurance, software and independent expert advisory services to different industries. A key challenge for DNV GL is to assure and verify systems and the use of systems which are based on data-driven models. In this thesis, we develop data-driven prediction models and anomaly detection models for applications in the maritime industry. Different modifications are proposed to improve existing anomaly detection techniques, and to improve their efficiency. We investigate different evaluation techniques, such as cross validation, to ensure robust performance on unseen future datasets. Since the performance of data-driven models rely on the experience in the training data, thorough dataset analysis is suggested. A novel training data-centric approach to explain and interpret data-driven models are proposed.

The doctoral rpoject concerns statistical sensor-based regression models and anomaly detection methods, evaluation methods, performance measures, and explainability and interpretation of black box models. The main contributions of the thesis include: - Development of a novel method to enhance the explanability and interpretations of opaque machine learning and statistical models. - Modifications of the semi-supervised anomaly detection framework based on reconstruction with AAKR and residuals analysis using SPRT to: 1) reduce computation time; 2) make it possible to distinguish between explanatory and response variables; and 3) use a regional credibility estimation in the residuals analysis to improve accuracy. - Demonstrations of the usefulness of data-driven methods for anomaly detection in maritime applications - Demonstrations of the importance of appropriate cross validation techniques for performance evaluation, especially for sequential sensor data.

Maintenance and inspections of ships performed by DNV GL have traditionally been based on a preventive scheme where components have been overhauled or maintained according to a time schedule. This philosophy is based on the assumption that a component has a defined lifetime, after which its failure rate increases. However, estimates of lifetime have large uncertainties and a large percentage of failures are not age-related, and are therefore not adequately addressed by preventive scheduled maintenance. DNV GL aim to develop radically new statistical approaches based on the recent availability of large arrays of sensors, which monitor condition and behaviour of machinery and power systems. Data are ubiquitous: almost every activity in which DNV GL engages produces and requires data. More and more ships get sensors installed, collecting more and more data. Data are the critical inputs into almost all decisions. Statistical inference is needed to turn data into knowledge, to understand unexplained mechanisms, discover hidden patterns, and predict future behaviours. As the accessibility, volume and complexity of data increases, new model based statistical and machine learning methods must be developed. Sensor data from ship generate very high dimensional time series, which relate to each other either in a known way (network of sensors) or in terms of stochastic dependence (correlation, coherence). These will be analysed for motif discovery, anomaly detection and classification. The purpose is to automatically alert about shifts in trends, variability or extremes, about changing patterns of behaviour or about any potential deviations from the norm. We will in particular consider state space models and other Bayesian hierarchical models. Both classification of faults, prediction of condition and faults, and optimization of performance are possible key aspects of this PhD.

Publications from Cristin

No publications found

No publications found

No publications found

No publications found

Funding scheme:

NAERINGSPH-Nærings-phd