PERIODO: OTTOBRE - NOVEMBRE 2016

The purpose of the course is to convey the main ideas behind Structural Health Monitoring (SHM), with a particular emphasis on the data-based approach (although some aspects of model-based SHM will be covered). The course will mainly concentrate on vibration-based SHM, but will also cover aspects of wave-based SHM including traditional non-destructive evaluation methods like Acoustic Emissions. The course will be largely self-contained in terms of its mathematical content, but will assume a good familiarity with the sort of mathematics covered in an undergraduate engineering programme e.g. matrix analysis, fourier analysis and linear differential equations.

Phase 1: Fundamentals of SHM. This will cover the basic material including motivation for SHM and the basic frameworks and methodologies. There will also be a lecture on typical sensor systems. (4 Lectures). Phase 2: Statistics and Machine Learning. This will cover the basic principles of statistical pattern recognition and their extensions into machine learning. Only a very basic knowledge of statistics will be assumed and will cover foundational material. The course will build up to the point where modern methods like Gaussian process regression and support vector classification are covered. (8 Lectures). Phase 3: Vibration-based and Wave-based SHM. This will cover the main features for the two types of SHM, including some of the basic and prerequisite vibration and wave theory. There will be illustrations from real experimental studies. (6 Lectures). Phase 4: Case Studies. This will cover two detailed case studies: one for vibration-based SHM and one for wave-based. (2 Lectures).