Abstract
An adaptive multi-tiered framework, that can be utilised for designing a context-aware cyber physical system to carry out smart data acquisition and processing, while minimising the amount of necessary human intervention is proposed and applied. The proposed framework is applied within the domain of offshore asset integrity assurance. The suggested approach segregates processing of the input stream into three distinct phases of Processing, Prediction and Anomaly detection. The Processing phase minimises the data volume and processing cost by analysing only inputs from easily obtainable sources using context identification techniques for finding anomalies in the acquired data. During the Prediction phase, future values of each of the gas turbine’s sensors are estimated using a linear regression model. The final step of the process— Anomaly Detection—classifies the significant discrepancies between the observed and predicted values to identify potential anomalies in the operation of the cyber physical system under monitoring and control. The evolving component of the framework is based on an Artificial Neural Network with error backpropagation. Adaptability is achieved through the combined use of machine learning and computational intelligence techniques. The proposed framework has the generality to be applied across a wide range of problem domains requiring processing, analysis and interpretation of data obtained from heterogeneous resources.
Highlights
There exists a growing demand for smart condition monitoring in engineering applications often achieved through evolution of the sensors used
An implementation of a context-aware cyber physical system using evolving inferential sensors for condition monitoring to predict the status of a gas turbine on an offshore installation has been successfully developed
A three-phase approach has been proposed: In the processing phase, historical data of 25 sensors was collected from different areas of turbine to train an evolving component (ANN-based) used as the basis of the prediction model
Summary
There exists a growing demand for smart condition monitoring in engineering applications often achieved through evolution of the sensors used This is especially true when some constraints are present that cannot be satisfied by human intervention with regard to decision making speed in life threatening situations (e.g. automatic collision systems, exploring hazardous environments, processing large volumes of data). Cyber physical systems (CPSes) integrate information processing, computation, sensing and networking, which allows physical entities to operate various processes in dynamic environments (Lee 2008) Many of these intelligent CPSes carry out smart data acquisition and processing that minimise the amount of necessary human intervention. Computational Intelligence (CI) techniques have been successfully applied to problems involving the automation of anomaly detection in the process of condition monitoring (Khan et al 2014) These techniques require training data to provide reliable and reasonably accurate specification of the context in which a CPS operates. The application of these novel approaches to developing evolving sensory systems for optimising the operation of an offshore gas turbine constitutes another original contribution of the paper that demonstrates practical benefits of the suggested methodology
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have