The Technology Of Pattern Identification Based On MAE For Damage Of The Plate Caused By Hypervelocity Impact

It has a history of nearly60years for the first of the the human space activities. The increase of space activities not only bring more convenient but also brings the deterioration of the space debris environment. Because of the space debris's high speed and high kinetic energy?its constitutes a great threat to operation of the spacecraft. In order to increase the spacecraft's survivability and reduce the probability of spacecraft safety accidents caused by space debris impact.It is necessary to research and development of the space debris hypervelocity impact of on-orbit perception system.Now the most commonly used technology in this perceptual system is modal acoustic emission(MAE) technology.The MAE technology linked the acoustic emission signal waveform and acoustic emission physical process produced by the impact together. It has been basically achieved on-orbit sense impact source localization module by using this technology.But there is little studied for the impact damage patten recognition. In addition there are some limitations in the hypervelocity impact damage pattern recognition: The premise of the identify is always single sensor and fixed position?and the identify metheds are always useing the statistical pattern classification methods.And there is some blindness in selecting characteristic parameters.Based on the background and current situation of the development above?we use the MAE technology research of the AL5A06aluminium alloy plate's impacting damage pattern recognition by AL2017aluminum alloy projectile.First? by using the method of wavelet decomposition and reconstruction separate the acoustic emission signal collected by sensors of each mode respectively. According to the characteristics of the signal separateing by wavelet decomposition and reconstruction?selected the characteristic parameters that can representative the modal information. And according to the characteristics of the planar acoustic emission?Proposed and verified a hypothesis about the produce process of tablet acoustic emission caused by hypervelocity impact: With the increase of impact velocity acoustic emission source change from the plate surface to the plate inside.Analyzed the relationship between the various modal characteristic parameters and the damage situation. By analyzing the extracted the modal characteristics of the relationship between the parameters and the impact velocity? propagation distance?we can see that: S0modal of each characteristic parameter values has a good linear relationship with the impact speed? with the impact speed increases. And because in the speed of2.5Km/s the target plate is punctured? the energy was taken away by the projectile so shifts in slowing phenomenon; A0mode of two linear relationship between the characteristic parameters with the impact speed is not good?and apparent minimum in the critical speed of2.5km/s when breakdown of the target plate; S2modal of each characteristic parameter values has a good linear relationship with the impact speed? and apparent minimum in the critical speed of2.5km/s when breakdown of the target plate. Found that S2modal significantly attenuation with the increase of the propagation distance attenuation?greatly influenced by the propagation distance; The relationship between the total energy with the impact velocity is not obvious?but the ratio of modal energy have good relationship with the impact speed.And the A0energy proportion decreases with the increase of impact velocity? the S0and the S2proportion of energy increases with the increase of impact velocity.For different phenomenon between the A0and S0energy ratio of sensor and distance?is due to A0attenuation rate with the increase of the propagation distance is greater than the total energy attenuation rate what is S0energy ratio on the contrary?and in the impact speed of2.5km/s appear different extreme value point is because when the projectile breakdown the plate?it take away a lot of energy what is predominantly S0modal energy.Then?for the blindness in selecting characteristic parameters of previous studies?Introduced evaluation algorithm—K-W test algorithm which is a algorithm for evaluation individual characteristic parameters's classification ability relative size for a specific pattern classification problems.And compared the recognition rate between the characteristic parameters which was screening by K-W test and all characteristic parameters which wasn't screening by K-W test. The results show that: Use all characteristic parameters recognition rate is morer than using the K-W test screening characteristic parameters? but the gap is not big.For reduce the complexity of the recognition and computational complexity?with the K-W test screening of characteristic parameters are necessary.Finally? aimed at the present situation of engineering that damage pattern recognition of the phenomenon of sensor utilization rate is not high?designs the multi-sensor data fusion algorithm for pattern recognition.Then from compared the average recognition rate of collision we can see that:The recognition accuracy by using multiple sensors is higher than the recognition accuracy by using inly single sensor?and5sensor's accuracy is higher than3sensor's accuracy.This paper's result and conclusions provides an important theoretical basis for the pattern recognition of hypervelocity impact injury?and with strong engineering value.