Modal learning displacement-strain transformation

The displacement-strain transformation is essential for revealing the internal mechanics of structures and developing strain measurement methods. This transformation typically depends on the environment of the structure. In contrast to the traditional invariant transformation, this paper proposes a method to obtain a variable displacement-strain transformation by self-learning of the modal parameters of the structure in operation. The beam experimental results demonstrate that the transformation is able to take account of different forms of excitation and to obtain strain measurements under sinusoidal and random excitation with up to 99.82% and 99.70% accuracy, respectively. Moreover, these results indicate that the proposed displacement-strain transformation is able to take account of the environmental conditions encountered in practical situations more consistently than conventional approaches. The introduction of a modal-learning displacement-strain transformation in the proposed approach provides a welcome boost to the development of strain measurement methods.