N-gamma discrimination method based on semi-supervised support vector machine

Abstract

The invention belongs to the technical field of neutron detection, and particularly relates to an n-gamma discrimination method based on a semi-supervised support vector machine, being characterized by comprising the following steps: a) performing analog-to-digital conversion on a detector output pulse signal by using an analog-to-digital converter, and collecting a digitized detector pulse signalto form a training data set; training a support vector machine by using the training data set in combination with a semi-supervised learning method to obtain an optimal classification hyperplane; b)digitalizing the newly detected pulse, and carrying out feature extraction preprocessing: extracting important feature information of the pulse signal, and c) inputting the extracted feature information into a support vector machine, and carrying out classification prediction on the newly detected pulse sample by using the classification hyperplane in combination with the extracted features.

Description

technical field [0001] The invention relates to the technical field of neutron detection, in particular to an n-γ discrimination method based on a semi-supervised support vector machine. Background technique [0002] Due to the inelastic scattering of neutrons and the surrounding environment, the radiation capture of slowed neutrons, etc., the occasions where neutrons exist are almost always accompanied by a large amount of γ-ray background; , Environmental radiation detection, medical and military fields are widely used, and n-γ discrimination (neutron and gamma ray discrimination) technology has gradually become a key technology in neutron detection. At present, the commonly used n-γ screening techniques include rise time method, charge comparison method, neural network method, pulse gradient method, and wavelet transform method; and these methods only use one or two kinds of characteristic information of the pulse signal for n-γ screening. . [0003] A typical detector ...

AI Technical Summary

Problems solved by technology

[0012] 5) For each ray particle measured by the detector, use S / S t As feature information, n-γ screening can be performed in combination with the statistical results in the above figure: for S / S t Less than γ accepts the boundary BoundaryG(BoundaryG=u n -3σ n , u n is the neutron distribution expectation, σ n neutron distribution standard deviation) is considered to be gamma rays, S / S t Greater than n accept boundary BoundaryN(BoundaryN=u γ +3σ γ ) are considered to be n-rays, and for S / S t Between BoundaryG and BoundaryN (such as image 3 middle shaded area) cannot be accurately identified

However, due to the amount of calculation and the calculation of complexity, this method only uses the sample points between 20 and 40 nS after the pulse peak as the feature information

At the same time, the neural network method has the following two disadvantages: one is that the method has a complex structure and a huge network; the other is that the method requires a large number of pulse signal data sets generated by known types of neutrons and gamma rays in the detector as standard samples. For training, it is almost impossible to obtain enough such standard samples in the actual environment

Images