Localization of hypervelocity impact-induced damage in thermal protection systems of aerospace vehicle using a weighted-KNN algorithm

Abstract

A large amount of micrometeoroids and orbital debris (MMOD) pose a high potential risk of impacts to the thermal protection system (TPS) of aerospace vehicles, threatening the structural integrity and safety. Therefore, real-time in-situ sensing of MMOD impacts is necessary. Due to the significant anisotropy and strong intermittency of the TPS, the traditional time difference of arrival (TDOA)-based method has limited accuracy. Targeting the localization of hypervelocity impact (HVI)-induced damage on the TPS, a weighted K-nearest neighbor (WKNN)-based method is proposed. Firstly, a TDOA vector database is constructed by repeating the low-velocity impact (LVI) experiments on the TPS specimen in the delineated zone to train the proposed method. Then, both LVI and HVI experiments are conducted to validate the proposed algorithm, and the effect of the K value and weighting method on the localization accuracy is analyzed. The results show that both LVI and HVI-induced damage on the TPS can be accurately localized using the proposed WKNN-based method. Adopting an appropriate K value can reduce the influence of database imbalance, and weighting the inter-vector distances significantly can improve the localization accuracy and robustness. The proposed method can provide valuable guidance for the on-orbit repair of TPS in practical engineering applications.