Detection method of loose parts in nuclear reactor based on eigenvector algorithm

A loose part in a nuclear reactor will generate an impact signal. Therefore, determining whether an impact signal exists in the detected signal is an effective method to ascertain whether loose parts are present in a nuclear reactor. In a practical environment, the impact signal may be submerged in background noise. A signal detection method based on the eigenvector algorithm (EVA) is proposed in this paper. Mathematical and statistical characteristic of kurtosis is particularly sensitive to the impact signal. Thus by the characteristics in EVA, extraction of the impact signal under strong noise can be realized. Simulation experiments are carried out to verify the proposed method. The results show that EVA is able to detect the impact signal.