Research On The Detection Technology Of Artificial Gas Leakage In Spacecraft Shell Structure

In low-Earth orbit(LEO),large,long-term in-orbit spacecraft are most likely to be hit by microdigres and orbital debris(MMOD),leading to degradation and deterioration of spacecraft st ructures and even gas leaks,posing a serious threat to the safety of spacecraft and ast ronauts.Sound emission detection tec hnology is conducive to timely in-orbit perception of whether gas leakage,and leak source positioning.However,due to th e gas leakage source signal has continuity,multi-characteristic f requency band,frequency dispersion characteristics and weak energy and other very special forms of expression,w ill not be able to according to the arrival t ime of the signal posit ioning,so that the spacecraft gas leakage to achieve rapid perception and accurate positioning has become an urgent problem to be solved.In order to solve this problem,a gas leakage sound source automatic positioning technology is proposed,using sensor array to locate the continuous leakage source in the spacecraft st ruct ure.At the same t ime,in order to assess the risk degree of leakage sou rce,neural network is used to ident ify the size of gas leak hole,which will minimize the risk of structural failure.The main contents are summarized below.Obtain the mechanism and attenuat ion characteristics of gas leakage sound emission signal.The gas leakage signal is collected by simulated gas leakage experiment,the leak signal is treated by Fourier transformation and numerical analysis method,and the time frequency characteristics,t ime domain amplitude and frequency domain energy attenuat ion characterist ics of different diameter leak leakage signals are analyzed.According to Lamb theory,the signal mode and its dispersion curve are calculated,the mode selected at posit ioning is determined,and the mechanism of modality is analyzed.Based on AR-AIC guidelines and chaos theory,the ident ificat ion of weak gas leakage signals is realized.The difference between the im pact signal and the leakage signal is that at the moment of achievement,after AR-AIC fitting,the impact signal w ill obtain the AIC minimum at the moment of achievement,while the AIC value of the leak signal is in a st raight line on the t imeline,according to the difference betw een the two AIC values to determine whether the leakage signal.In view of the recognit ion of weak leakage signal under the background of noise,the weak leakage signal is added to the chaot ic system,and the phase trajectory is changed from chaotic state to large periodic state to ident ify w hether leakage occurs.An automat ic positioning algorithm for gas leak sound sources w ith automatic window narrow band filtering is proposed.According t o the leakage signal multi-band characterist ics,frequency dispersion phenomenon and other special manifestations,through the design of automatic window narrow band filtering independent select ion of signal mult iple characterist ic f requency bands,and automatic frequency band weight ing and wave velocity compensation,not only to achieve the aluminum alloy plat e and reinforced plat e st ructure of different diameter leakage sound source accurate posit ioning,but also improve the applicability of the algorithm.The neural network is used to identify the size of t he leak hole.It is found t hat the average amplitude of gas leakage signal increases signif icantly w ith the increase of leakage hole,and the energy of each characteristic band decays with the distance of propagat ion,and the average amplitude and the energy of each band are select ed as training samples,and the ident ification of leak hole size is realized by establishing BP neural network.The research result s are of great pract ical significance to perfect the leak identif icat ion module and its engineering application of space debris in-orbit sensing system.