Localized damage identification in plate-like structures using self-powered sensor data: A pattern recognition strategy

This study presents a novel strategy for localized damage identification in plate-like structures based on the binary data provided by self-powered wireless sensors that use a pulse switching communication architecture. The energy-aware pulse switching communication architecture uses single pulses instead of multi-bit packets for information delivery resulting in discrete binary data. A system employing such an energy-efficient technology requires dealing with power budgets for sensing and communication that leads to unique time delay constraints. This paper presents a new paradigm for localized damage detection using time-delayed and sometimes limited binary data. The binary data from the sensor nodes resembles an image, or pattern. Data analysis using pattern recognition (PR) framework incorporating PR methods and a conditional probability chain is thus presented. Numerous features extracted from cumulative acceleration on dynamically loaded plates are used to determine damage indication parameters. Performance of the proposed damage detection strategy was evaluated through finite element simulations for the case of a simply supported aluminum plate under distributed harmonic loading. Different damage states were considered to calibrate the damage detection algorithm, and an uncertainty analysis was performed by superposing random noise to the input vectors. Results show that the proposed PR framework is capable of identifying damage even in the presence of high noise levels. The findings demonstrate satisfactory performance of the localized damage detection strategy, and the applicability of PR to detect and localize damage in plate-like structures from the time-delayed binary data generated by self-powered sensors.