The outlook regarding instalments of wind energy infrastructures is rapidly and steadily increasing over recent and forthcoming years. Big markets such as China had a share of 48.6% of global wind installations in 2015 (92,981 wind turbines in service by then). Moreover, Europe have plans to use wind energy to produce 24.4% of EU's electricity demand with 66GW offshore and 254 GW onshore wind. This will essentially increase the importance of structural health monitoring (SHM) to reduce maintenance and service costs, which should consequently reduce the numbers of unexpected structural failures and save costs in repairs and inspections.
Currently, passive techniques for SHM are being intensively researched to ensure the service time of wind turbines run for the anticipated lifetime of such structures. Statistical analysis of data collected from acoustic emissions (AE) testing is the main area of focus for this research. This technique is used broadly, but the influencing parameters and factors collected and derived from this technique still needs to be investigated to confirm its credibility in online SHM. Hence, the scope of this PhD will be to implement statistical pattern recognition tackle existing issues. Nonparametric and parametric statistical methods will be studied to discern AE signals from noise, increase the accuracy of time-of-flight of signals, and subsequently provide a sophisticated approach to acquiring AE signals other than using the traditional threshold-based technique. Data acquired from field tests and laboratory experiments will be used to validate this research.