Pulse Signal Processing Method Based On Wavelet Transform

There is a complex radiation environment in the terrestrial space,causing damage to the the spacecraft and astronauts.In this paper,based on the signal caused by the radiation particles detected by the plastic scintillator,the wavelet transform is used to process the signal.In the space radiation environment,the interference of the data collected by the detector mainly includes noise interference and false event interference.There are many processing methods for pulse signal,such as analog signal processing and digital signal processing,among which digital signal process is a good choice for signal processing due to its relatively simple programming and convenient algorithm modification.In this paper,three methods are used to filter the pulse signal to calculate their FOM values.Among the three methods,which are smooth denoising method,discrete Fourier denoising method and wavelet denoising method can improve the FOM of the original data by1.6%,5.1% and 12.1% respectively.Wavelet denoising method has the best improvement effect on FOM.The experiment of detector test,data acquisition and interference process was carried out with D-T neutron generator.When analyzing influence of noise,the value of FOM is 0.724 for neutrons and gamma rays with the equivalent electronic energy of more than 150 ke V generated by D-T neutron generator after analog signal amplification without digital signal processing.When different wavelet functions are used to process pulse signals,db20 wavelet has the best improvement on particle discrimination,and FOM increases from 0.724 to 0.832,with an increase of 14.92%.The optimization effect of Db4 wavelet,db8 wavelet and db12 wavelet is similar to that of db20 wavelet,but higher than that of haar wavelet.Haar wavelet has the worst ability to improve particle identification due to its discontinuous characteristics.FOM increases from 0.724 to 0.744,with an increase of only 2.76%.As particle energy increases,the improvement effect of wavelet denoising ability for neutron and gamma ray begins to decrease.At this time,the signal distortion caused by haar wavelet processing is worse,and even reduces the neutron/gamma discrimination ability of the detector.Then the modular maximum method based on wavelet transform is used to identify the pile-up signal.After the impulse signal is transformed by wavelet transform on several binary scales,the peak distribution of the pulse waveform can be judged by comparing the distribution of the maximum value of the wavelet system.After processing,the pile-up event recognition algorithm based on wave crest removes the heaping event which accounts for 12.1% of the total number of events,and increases FOM from 0.724 to 0.728.However,the pile-up event recognition algorithm based on wavelet transform removes 10.2% of the total number of events,and increases FOM from 0.724 to0.734.The results show that the recognition algorithm based on wavelet transform can remove less pile-up events and improve FOM better,identifying pile-up events more accurately.