Research On The Signal Processing Technology Of High Count Rates Digitized Spectrometer

Coal is the primary fuel-fired in power plants. Accurate on-line measurement of the composition of coal can guide the optimization combustion of boilers, improve boilers'combustion efficiency, reduce pollutant emissions and reduce adverse incidents. Prompt gamma neutron activation analysis (PGNAA) is one of the main methods which can achieve the on-line measurement of coal and other bulk materials. High count rates digitized multi-channel gamma spectrometry analyzer is the core components which can guarantee the high precision on-line measurement of material composition. Theoretical analysis and experimental research about the key technologies to acquire and process high count rate pulse signal, to build high-resolution gamma spectrometry were done in this paper.The test platform was designed firstly and large size crystal detectors and high-speed ADC were chosen. While doing on-line measurement of coal, more gamma signal could be collected per unite time. Signal analysis platform was designed to achieve filtering algorithms of digitized pulse signal and built gamma spectroscopy. By analyzing gamma spectroscopy, signal processing algorithms which presented in this paper could be verified and evaluated.The characteristics of gamma signal pulses obtained from scintillator crystal detector were analyzed and the characteristics of a variety of filtering algorithms were analyzed in this paper. And based on the above analysis, Least squares filtering method, Butterworth filtering, Chebyshev type I filtering, Chebyshev type II filtering and Elliptic filtering were tested. The experiment showed:third-order Butterworth low-pass filtering was most suitable for signal's processing and the optimum filtering frequency was 1.5MHz.The process of establishing gamma spectrometry was studied and for the pulse signal accumulation effect, four pulse peak searching algorithms and four overlapping peak processing algorithms were raised by analyzing the characteristics of pulse signal. After building gamma spectrometry and analyzing gamma spectrometry on signal analysis platform, the experiment showed:derivative peak search method and multi-point comparison peak search method were most suitable for peak's searching. Deleting method and model approach method were most suitable for overlapping peaks processing.The online measurement experiment for coal showed that the performance of the high count rate digital spectrometer was better than the performance of traditional electronics spectrometer based on small and medium size crustal detector obviously.