The Recovery Method Research Of Acoustic Emission Source Signal Of Refractory Materials Based On Particle Filter

The main role of refractories is structural materials and linings which resist the high temperature in a variety of thermal equipment and high temperature container and it is significant to guarantee the safe operation of high-temperature equipment.Refractory damage which is affected by the actual working conditions and the environment seriously endanger normal operation even the quality of the furnace products.Therefore refractory damage identification is important.Refractory damage identification based on the acoustic emission signal analysis has become a hotspot of current research,but subject to impact of the delivery channel,distortion AE source information directly affect the accuracy of damage identification.So this paper carried out research of AE source signal recovery,the main contents are as follows.a.Considering the characteristics that the acoustic emission system contains both non-linear situation and linear situation,this paper proposed the source signal restored method based on improved Rao-Blackwellized particle filter(RBPF)which estimate system state(source signal)by combining state-space model,The paper demonstrated the performance and advantage of improved RBPF on the recovery of refractory acoustic emission signals by theoretical and experimental angles.b.In order to establish an appropriate acoustic emission system model and because of the similarity of acoustic emission signal and seismic wave signal,the paper established acoustic emission system state space model by the introduction of seismic wave model and modifying the parameters,then restored actual damage acoustic emission signal source signal by using standard PF,RBPF and improved RBPF,the three recovery signal were assigned weights according to the simulation and obtained the signal closest to the true value,it found that improved RBPF algorithm are the best when obtained signals are compared with the three algorithms recovery results.c.The paper analyzed the power spectrum of acquisition signal and source signalin order to further illustrate the effect of the source signal recovery.In both cases,matrix cracking clocked components were 6~8kHz and5~9kHz,matrix crack extension clocked components were 40~60kHz and 10~20kHz,the recovery effect of improved RBPF are more representative through previous analysis,clocked components of matrix cracking and matrix crack extension should be 5~9kHz and 10~20kHz after using the algorithm.The signal restored method based on improved RBPF which was proposed by this paper not only restored refractory AE source signal effectively but also provided more accurate basis for recognition of their injury patterns,it provides a new idea for the study of damage material.