A Research On The Metal Collision Weak Signal Detection Under The Strong Background Noise

Nuclear power has become an important power for its stability and efficient production and has a great prospect,but there also exists security risks.In a reactor,the loose parts may cause serious accidents.Signals for loose parts should be monitored so that accidents can be effectively avoided or eliminated.Based on the study on “The research of loose parts monitoring system in nuclear power station”,this paper studies deeply to the signal caused by the loose parts in the reactor.How to detect the impulse signal under the strong background noise is the key content of this paper.The main contents are as follows:Due to the special operation of the nuclear power plant,it is very difficult to collect the actual signal in the reactor.According to common practice,the plate or surface experiment to get the impact signal is usually adopted.So we can use the impact signal to analog the signals of the loose parts.In this paper,we have designed the scheme of the surface impact experiment and boiler experiment to get the experimental data,then analyzed it and provided data support for the research of new theoretical.This paper analyzes the characteristics of the impact signal caused by loose parts in the reactor from the traditional signal detection method such as correlation detection,sampling integral.In the time-frequency domain,wavelet transform and wavelet energy spectrum are performed for in-depth theoretical study and simulation experiment.From the experiment results,we can get the characteristics of the signal so as to provide a good and real database for the research of subsequent algorithm.According to the characteristics of the impact signal of the loose parts and nonlinear theory,this paper proposes the nonlinear theory detection method based on wavelet transform.Matlab simulation experiment results show that this method can detect the impact signal effectively even at very low SNR.However,this method cannot extract the characteristics of the impact signal,such as the features of time domain and frequency domain.In order to better analyze the characteristics of the impact signal and strong background noise and provide a new idea for the development of the LPMS.This paper presents a weak signal detection algorithm of mental collision based on independent component analysis method.Matlab simulation experiments show that this method can effectively isolate the impact signal and background noise,and the signal can maintain original feature after separation.The separation result is relatively good,thus it can provide a new algorithm idea for the resolution of key issues in the development of LPMS.