Detecting structural damage in civil structures through non-destructive means is a growing field in civil engineering. There are many viable methods, but they can often be time consuming and costly; requiring large amounts of data to be collected. By determining which data are the most optimal at detecting damage and which are not the methods can be better optimized. The objective of this thesis was to adapt an existing method of data optimization, used for damage detection in mechanical engineering applications, for use with civil structures. The existing method creates Parameter Signatures based on characteristics from the system being analyzed, from which preferred locations for recording data are determined. For civil structures this method could potentially be used to locate the preferred locations to place accelerometers such that the minimum number of accelerometers is needed to properly detect the location and severity of damage in the structure. This method was first tested on fully analytical computer model structures under perfect conditions to determine its mathematical feasibility with civil structures. It was then tested on data recorded from physical test structures under “real-world” conditions to determine its feasibility as an actual damage detection optimization procedure. Results from the analytical testing show that this is in fact a viable method for determining the preferred sensor positions in civil structures. Furthermore, these results were verified for a variety of excitation types. Physical testing was inconclusive, leading to great insight about what obstacles are impeding this method and should looked at in future research.

A method of sensor location selection is introduced for distributed parameter systems. In this method, the sensitivities of spatial outputs to model parameters are computed by a model and transformed via continuous wavelet transforms into the time-scale domain to characterize the shape attributes of output sensitivities and accentuate their differences. Regions are then sought in the time-scale plane wherein the wavelet coefficient of an output-sensitivity surpasses all the others’ as indication of the output sensitivity’s uniqueness. This method yields a comprehensive account of identifiability each output provides for the model parameters as the basis of output selection. This output selection strategy is evaluated for a numerical case of pollutant dispersion by advection and discussion in a two-dimensional area.