Data analysis system and method based on semiconductor detector

Abstract

The invention discloses a data analysis system and method based on a semiconductor detector, and relates to the technical field of data analysis, and the analysis method comprises the following steps: constructing a detection data management platform to receive detection data and environment data collected by the semiconductor detector, and generating a corresponding detection record; carrying out multi-dimensional evaluation on the detection data stored in any detection record and carrying out anomaly identification; feature extraction is carried out on the detection data of each dimension, and radiation types existing in detection records are identified; carrying out difference analysis on the detection records of any abnormal identification, and extracting abnormal factors influencing various radiation types; setting corresponding anomaly evaluation thresholds for different radiation types; generating a real-time detection record for the semiconductor detector in real time, and performing anomaly evaluation on the real-time detection record; and calling a corresponding abnormal evaluation threshold value from the real-time detection record, and judging whether risk early warning is carried out or not.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to a data analysis system and method based on semiconductor detectors. Background Art

[0002] With the continuous advancement of science and technology, semiconductor detectors, as core components of high-sensitivity radiation detection, convert particle energy into electrical signals through the ionization effect. The reliability of the output data directly determines the overall performance of the detection system. Due to their advantages such as high sensitivity, high resolution and fast response, semiconductor detectors are widely used in many fields and have become an important part of modern detection technology.

[0003] Traditional anomaly detection methods usually rely on a set fixed threshold. When the detected signal exceeds the threshold, the system will judge it as an anomaly. However, in actual radiation monitoring, the background radiation level may fluctuate with time and environmental conditions. This makes the fixed threshold anomaly identification method difficult to adapt to actual conditions, easily leading to false alarms and missed alarms, affecting the accuracy and reliability of monitoring. Summary of the Invention

[0004] The object of the present invention is to provide a data analysis system and method based on semiconductor detectors to solve the problems raised in the prior art.

[0005] To achieve the above object, the present invention provides the following technical solution: a data analysis method based on a semiconductor detector, the analysis method comprising the following steps:

[0006] Step S100: Build a detection data management platform to receive detection data and environmental data collected by each detection process of the semiconductor detector and generate corresponding detection records; conduct multi-dimensional evaluation of the detection data stored in any detection record and identify anomalies in any detection record;

[0007] Step S200: Analyze the detection data corresponding to each dimension in any detection record, extract features from the detection data of each dimension; combine the features extracted from the detection record to identify the radiation type present in the detection record;

[0008] Step S300: performing a difference analysis on any detected abnormality detection records to extract abnormality factors affecting various radiation types; setting corresponding abnormality assessment thresholds for different radiation types based on the abnormality factors contained in different detection records;

[0009] Step S400: generating a real-time detection record for the data collected in real time by the semiconductor detector, and performing an abnormality assessment on the real-time detection record; retrieving a corresponding abnormality assessment threshold for the real-time detection record, and determining whether to issue a risk warning for the real-time detection record.

[0010] Furthermore, step S100 includes the following steps:

[0011] Step S101: Several environmental sensors are pre-installed on the semiconductor detector. Whenever the semiconductor detector is used to detect a desired detection area, the data collected by the semiconductor detector is set as detection data, and the data collected by the several environmental sensors is set as environmental data. The detection data and environmental data are synchronously transmitted to the detection data management platform. The desired detection area is a target area pre-planned before each use of the detector.

[0012] Step S102: Summarize the detection data and environmental data received in the detection data management platform to generate a detection record; extract the detection data in any selected detection record, divide the detection data into several dimensions according to the data type, preset corresponding evaluation rules for each dimension, and obtain the evaluation value of the detection data in any dimension; the data type of the detection data includes various indicators such as radiation intensity and radiation range for identifying whether there are abnormalities in the detection area, and set corresponding evaluation rules for each indicator. By comparing the deviation between each indicator and the preset normal value range, the detection result of the detection record is evaluated for abnormalities;

[0013] Step S103: Calculate the average value of the evaluation values under each dimension in the selected detection record to obtain the evaluation value of the selected detection record as P; preset an evaluation threshold P th , if P<P th , the selected detection record is set as an abnormal record, otherwise, the selected detection record is set as a normal record; the detection record is identified as abnormal through a preset fixed evaluation threshold, and the abnormal data presented by the abnormal record can be accurately captured by analyzing the abnormal data in the subsequent feature extraction process.

[0014] Furthermore, step S200 includes the following steps:

[0015] Step S201: arbitrarily select a detection record, and arbitrarily select detection data under a dimension from the selected detection record to obtain the data range under the selected dimension, wherein the data range of the i-th dimension is set to (a i ,b i ); extract the preset evaluation rules of the i-th dimension and obtain the expected data range of the i-th dimension (a ex ,b ex), if (a i ,b i )∈(a ex ,b ex ), the i-th dimension in the selected detection record is set as the normal dimension, otherwise, the i-th dimension is set as the abnormal dimension;

[0016] Step S202: Randomly select an abnormal record. If the i-th dimension in the selected abnormal record is an abnormal dimension, feature extraction is performed from the detection data of the i-th dimension. The features of all abnormal dimensions in the selected abnormal record are summarized to obtain a feature set of the selected abnormal record. Feature extraction is not required for abnormal dimensions that exist in normal records because they do not affect the actual detection process. However, abnormal dimensions in abnormal records actually affect the detection process and require feature extraction.

[0017] Step S203: Pre-establish a radiation type database, storing a plurality of radiation types and a plurality of radiation feature sets in the database, wherein any radiation type matches a radiation feature set; compare the feature set of the selected abnormal record with any radiation feature set; if any radiation feature in a radiation feature set is identical to a feature in the feature set, set the radiation type matched by the radiation feature set to be compared as a radiation type of the selected abnormal record, thereby obtaining a plurality of radiation types of the selected abnormal record;

[0018] Step S204: Assume that any normal record contains several features, then compare the features with the radiation features in the radiation type database. If several radiation types are obtained after comparing the features of the normal record, then arbitrarily select one of the radiation types and extract all abnormal records containing the selected radiation type. Assume that the dimension where there is a radiation feature in the selected radiation type is the i-th dimension, and arbitrarily select an abnormal record from all the extracted abnormal records. The data range of the i-th dimension in the selected abnormal record is (a i ,b i ) and the expected data range is (a ex ,b ex ), the deviation amplitude of the i-th dimension in the selected abnormal records is f i =(b i -b ex ) / b ex ;

[0019] Step S205: Obtain the deviation amplitude of the i-th dimension in the remaining abnormal records to obtain the abnormal deviation amplitude range of the i-th dimension ((f i ) min ,(f i )max ); If there is a normal record containing the selected radiation type, the deviation amplitude of the i-th dimension is f1, and f1∈((f i ) min ,(f i ) max ), then the abnormal deviation amplitude range of the i-th dimension is corrected to (f1, (f i ) max );

[0020] Step S206: Randomly select a radiation type. If there is a normal record containing the selected radiation type, then merge the selected radiation type and the corresponding abnormal deviation amplitude range in the abnormal record containing the selected radiation type to generate a data set of the selected radiation type. If there is no normal record containing the selected radiation type, then only generate the selected radiation type in the abnormal record containing the selected radiation type, and obtain a set of radiation types existing in any detection record.

[0021] If a radiation type only exists in the abnormal record, it means that when the corresponding radiation characteristics deviate, it will directly lead to the occurrence of an abnormal situation. If it appears in both detection records at the same time, it means that the abnormality caused by the radiation type has a degree, and only when it exceeds this degree will it lead to the occurrence of an abnormal situation. Therefore, it is also necessary to extract the degree of the radiation type, that is, to correct the range of the abnormal deviation.

[0022] Furthermore, step S300 includes the following steps:

[0023] Step S301: The environmental data recorded in any detection record is divided into several categories of environmental data according to data type, and the corresponding environmental factors are extracted for each category of environmental data; an environmental factor is randomly selected, and the data range of the selected environmental factor in each normal record is obtained, and a union operation is performed to obtain the normal data range of the selected environmental factor;

[0024] Step S302: Randomly select an abnormal record and obtain the data range (r1, r2) of the jth environmental factor, and set the normal data range of the jth environmental factor as R j ,like Then set the jth environmental factor as an abnormal factor to select abnormal records, and obtain the normal data range R j The minimum value (r j ) min and the maximum value (r j ) max , according to the formula:

[0025]

[0026] Among them, Max() is the maximum value function; the deviation amplitude η of the jth environmental factor in the selected abnormal record is calculated j ;

[0027] Step S303: arbitrarily select two abnormal records containing the same radiation type, arbitrarily select a radiation feature from the same radiation type, and obtain the deviation amplitude of the two abnormal records in the abnormal dimension where the selected radiation feature is located. Assuming that the selected radiation feature is in the i-th abnormal dimension, the deviation amplitude difference Δf between the two abnormal records is obtained. i ;

[0028] Step S304: Obtain each abnormal factor in the two abnormal records respectively, and obtain the deviation amplitude difference of the environmental factors corresponding to any abnormal factor between the two abnormal records, wherein the deviation amplitude difference of the jth environmental factor is set to Δη j , according to the formula:

[0029]

[0030] Where m is the total number of environmental factors; the deviation influence coefficient y of the jth environmental factor is calculated j ; Obtain the deviation amplitude difference between the two abnormal records of each radiation feature in the same radiation type, and obtain the deviation influence coefficient of the j-th environmental factor on each radiation feature;

[0031] Step S305: Obtain the deviation influence coefficient of any environmental factor on any radiation feature between any two abnormal records, calculate the average value to obtain the average influence coefficient of any environmental factor on any radiation feature; set the average influence coefficient of the jth environmental factor on the i2th radiation feature as X (j,i2) ; Randomly select an abnormal record and obtain the deviation amplitude of the jth environmental factor as η j The deviation from the i1th radiation feature in the hth radiation type is (f h ) i1 , according to the formula:

[0032]

[0033] Among them, i1, i2 and j are all positive integers, and i1∈(1,n), i2∈(1,n), j∈(1,m), n is the number of radiation features matched by the h-th radiation type, and m is the total number of environmental factors; the anomaly assessment value YP of the h-th radiation type in the selected anomaly record is calculated h ; Obtain the abnormal evaluation value of the hth radiation type in each abnormal record, and select the abnormal evaluation value with the smallest value as the abnormal evaluation threshold of the hth radiation type;

[0034] By setting an abnormality assessment threshold for each radiation type, the abnormal factors in the detection area are subdivided. Because different detection areas cannot be exactly the same, using a fixed assessment threshold will definitely cause deviations in different detection areas, and the greater the difference, the greater the deviation. By subdividing the abnormal factors, this problem can be effectively solved.

[0035] Furthermore, step S400 includes the following steps:

[0036] Step S401: Acquire detection data and environmental data collected in real time by the current semiconductor detector, obtain features and abnormal factors contained in the real-time detection record; compare the contained features with a radiation feature set in a radiation type database to obtain several radiation types of the real-time detection record;

[0037] Step S402: Obtain the deviation amplitude of any feature and the deviation amplitude of any abnormal factor in the various radiation types recorded in the real-time detection, extract the average influence coefficient of any abnormal factor on any feature, and calculate the abnormal evaluation value YP of the real-time detection record. now ;

[0038] Step S403: Obtain anomaly assessment thresholds of any radiation type in the real-time detection record, and sum them up to obtain an expected anomaly assessment threshold YP th , if YP now >YP th , then a risk warning is issued for the real-time detection record.

[0039] In order to better implement the above method, a data analysis system based on semiconductor detectors is also proposed. The analysis system includes a historical detection analysis module, a radiation feature analysis module, an abnormality threshold division module, and a risk abnormality assessment module.

[0040] The historical detection analysis module is used to build a detection data management platform to receive the detection data and environmental data collected by the semiconductor detector during each detection process and generate corresponding detection records; conduct multi-dimensional evaluation of the detection data stored in any detection record and identify anomalies in any detection record;

[0041] The radiation feature analysis module is used to analyze the detection data corresponding to each dimension in any detection record and extract features from the detection data of each dimension; combine the various features extracted from the detection record and identify the radiation type present in the detection record;

[0042] The abnormality threshold classification module is used to perform differential analysis on any detection records that identify abnormalities and extract abnormal factors that affect various radiation types; based on the abnormal factors contained in different detection records, corresponding abnormality assessment thresholds are set for different radiation types;

[0043] The risk anomaly assessment module is used to generate a real-time detection record for the data collected by the semiconductor detector in real time, and perform an anomaly assessment on the real-time detection record; retrieve the corresponding anomaly assessment threshold for the real-time detection record, and determine whether to issue a risk warning for the real-time detection record.

[0044] Furthermore, the historical detection and analysis module includes a detection record collection unit and an abnormal record identification unit;

[0045] The detection record acquisition unit is used to build a detection data management platform to receive the detection data and environmental data collected by the semiconductor detector during each detection process and generate corresponding detection records; the abnormal record identification unit is used to carry out multi-dimensional evaluation of the detection data stored in any detection record and identify abnormalities in any detection record.

[0046] Furthermore, the radiation signature analysis module includes a radiation signature extraction unit and a radiation type identification unit;

[0047] The radiation feature extraction unit is used to analyze the detection data corresponding to each dimension in any detection record and extract features from the detection data of each dimension; the radiation type identification unit is used to combine the various features extracted from the detection record and identify the radiation type existing in the detection record.

[0048] Furthermore, the abnormal threshold division module includes an environmental factor extraction unit and a dynamic threshold adjustment unit;

[0049] The environmental factor extraction unit is used to perform differential analysis on any detection records that identify abnormalities and extract abnormal factors that affect various radiation types; the dynamic threshold adjustment unit is used to set corresponding abnormality assessment thresholds for different radiation types based on the abnormal factors contained in different detection records.

[0050] Furthermore, the risk anomaly assessment module includes a real-time risk assessment unit and an abnormal risk early warning unit;

[0051] The real-time risk assessment unit is used to generate a real-time detection record for the data collected by the semiconductor detector in real time and perform an abnormality assessment on the real-time detection record; the abnormal risk warning unit is used to retrieve the corresponding abnormality assessment threshold for the real-time detection record and determine whether to issue a risk warning for the real-time detection record.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] 1. This invention optimizes the traditional abnormality monitoring method that sets a fixed threshold, dynamically adjusting the abnormality assessment threshold for different radiation environments, helping workers accurately identify abnormal radiation conditions in different environments, and significantly improving the accuracy of semiconductor detector monitoring;

[0054] 2. The present invention identifies the radiation types present in the detection records, sets a corresponding abnormality judgment method for each radiation type, and accumulates the abnormality assessment thresholds of various radiation types present in each detection record to obtain the actual assessment threshold of the detection record. This can accurately assess various environmental conditions, help staff discover abnormalities in a timely manner, and ensure the safe implementation of the detection process;

[0055] 3. The present invention analyzes the impact of environmental factors on various radiation types and identifies the actual impact of radiation types in different environments. Compared with conventional fixed threshold comparisons, it can more accurately identify anomalies, reduce the possibility of missed detections, and ensure the reliability of detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A schematic diagram of the steps of a data analysis method based on a semiconductor detector;

[0057] Figure 2 This is a structural diagram of a data analysis system based on semiconductor detectors. DETAILED DESCRIPTION

[0058] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0059] Example: Figures 1 to 2 As shown, the present invention provides a data analysis method based on a semiconductor detector, the analysis method comprising the following steps:

[0060] Step S100: Build a detection data management platform to receive detection data and environmental data collected by each detection process of the semiconductor detector and generate corresponding detection records; conduct multi-dimensional evaluation of the detection data stored in any detection record and identify anomalies in any detection record;

[0061] Wherein, step S100 includes the following steps:

[0062] Step S101: Several environmental sensors are pre-installed on the semiconductor detector. Whenever the semiconductor detector is used to detect the desired detection area, the data collected by the semiconductor detector is set as detection data, and the data collected by the several environmental sensors is set as environmental data. The detection data and environmental data are synchronously transmitted to the detection data management platform;

[0063] Step S102: Summarize the probe data and environmental data received from the probe data management platform to generate a probe record; extract the probe data from any selected probe record, divide the probe data into several dimensions according to data type, preset corresponding evaluation rules for each dimension, and obtain the evaluation value of the probe data in any dimension;

[0064] Step S103: Calculate the average value of the evaluation values under each dimension in the selected detection record to obtain the evaluation value of the selected detection record as P; preset an evaluation threshold P th , if P<P th , the selected detection record is set as an abnormal record, otherwise, the selected detection record is set as a normal record.

[0065] Step S200: Analyze the detection data corresponding to each dimension in any detection record, extract features from the detection data of each dimension; combine the features extracted from the detection record to identify the radiation type present in the detection record;

[0066] Step S200 includes the following steps:

[0067] Step S201: arbitrarily select a detection record, and arbitrarily select detection data under a dimension from the selected detection record to obtain the data range under the selected dimension, wherein the data range of the i-th dimension is set to (a i ,b i ); extract the preset evaluation rules of the i-th dimension and obtain the expected data range of the i-th dimension (a ex ,b ex ), if (a i ,b i )∈(a ex ,b ex ), the i-th dimension in the selected detection record is set as the normal dimension, otherwise, the i-th dimension is set as the abnormal dimension;

[0068] Step S202: arbitrarily select an abnormal record. If the i-th dimension in the selected abnormal record is an abnormal dimension, extract features from the detection data of the i-th dimension; summarize the features of all abnormal dimensions in the selected abnormal record to obtain a feature set of the selected abnormal record;

[0069] Step S203: Pre-establish a radiation type database, storing a plurality of radiation types and a plurality of radiation feature sets in the database, wherein any radiation type matches a radiation feature set; compare the feature set of the selected abnormal record with any radiation feature set; if any radiation feature in a radiation feature set is identical to a feature in the feature set, set the radiation type matched by the radiation feature set to be compared as a radiation type of the selected abnormal record, thereby obtaining a plurality of radiation types of the selected abnormal record;

[0070] Step S204: Assume that any normal record contains several features, then compare the features with the radiation features in the radiation type database. If several radiation types are obtained after comparing the features of the normal record, then arbitrarily select one of the radiation types and extract all abnormal records containing the selected radiation type. Assume that the dimension where there is a radiation feature in the selected radiation type is the i-th dimension, and arbitrarily select an abnormal record from all the extracted abnormal records. The data range of the i-th dimension in the selected abnormal record is (a i ,b i ) and the expected data range is (a ex ,b ex ), the deviation amplitude of the i-th dimension in the selected abnormal records is f i =(b i -b ex ) / b ex ;

[0071] Step S205: Obtain the deviation amplitude of the i-th dimension in the remaining abnormal records to obtain the abnormal deviation amplitude range of the i-th dimension ((f i ) min ,(f i ) max ); If there is a normal record containing the selected radiation type, the deviation amplitude of the i-th dimension is f1, and f1∈((f i ) min ,(f i ) max ), then the abnormal deviation amplitude range of the i-th dimension is corrected to (f1, (f i ) max );

[0072] Step S206: Randomly select a radiation type. If there is a normal record containing the selected radiation type, then merge the selected radiation type and the corresponding abnormal deviation amplitude range in the abnormal record containing the selected radiation type to generate a data set of the selected radiation type. If there is no normal record containing the selected radiation type, then only generate the selected radiation type in the abnormal record containing the selected radiation type, and obtain a set of radiation types existing in any detection record.

[0073] Example 1: arbitrarily select the i-th dimension, and obtain the abnormal deviation amplitude range of the i-th dimension in each abnormal record as (10%, 20%). The i-th dimension has a deviation amplitude of 15% in the normal record, so the abnormal deviation amplitude range needs to be corrected to (15%, 20%). Arbitrarily select a radiation type, and set the radiation type to only include the radiation characteristics corresponding to the i-th dimension. If the normal record contains the radiation type, it is necessary to merge the radiation type and the abnormal deviation amplitude range (15%, 20%) to generate a data set. If it is not contained in the normal record, only the radiation type needs to be recorded.

[0074] Step S300: performing a difference analysis on any detected abnormality detection records to extract abnormality factors affecting various radiation types; setting corresponding abnormality assessment thresholds for different radiation types based on the abnormality factors contained in different detection records;

[0075] Wherein, step S300 includes the following steps:

[0076] Step S301: The environmental data recorded in any detection record is divided into several categories of environmental data according to data type, and the corresponding environmental factors are extracted for each category of environmental data; an environmental factor is randomly selected, and the data range of the selected environmental factor in each normal record is obtained, and a union operation is performed to obtain the normal data range of the selected environmental factor;

[0077] Step S302: Randomly select an abnormal record and obtain the data range (r1, r2) of the jth environmental factor, and set the normal data range of the jth environmental factor as R j ,like Then set the jth environmental factor as an abnormal factor to select abnormal records, and obtain the normal data range R j The minimum value (r j ) min and the maximum value (r j ) max , according to the formula:

[0078]

[0079] Among them, Max() is the maximum value function; the deviation amplitude η of the jth environmental factor in the selected abnormal record is calculated j ;

[0080] Step S303: arbitrarily select two abnormal records containing the same radiation type, arbitrarily select a radiation feature from the same radiation type, and obtain the deviation amplitude of the two abnormal records in the abnormal dimension where the selected radiation feature is located. Assuming that the selected radiation feature is in the i-th abnormal dimension, the deviation amplitude difference Δf between the two abnormal records is obtained. i ;

[0081] Step S304: Obtain each abnormal factor in the two abnormal records respectively, and obtain the deviation amplitude difference of the environmental factors corresponding to any abnormal factor between the two abnormal records, wherein the deviation amplitude difference of the jth environmental factor is set to Δη j , according to the formula:

[0082]

[0083] Where m is the total number of environmental factors; the deviation influence coefficient y of the jth environmental factor is calculated j ; Obtain the deviation amplitude difference between the two abnormal records of each radiation feature in the same radiation type, and obtain the deviation influence coefficient of the j-th environmental factor on each radiation feature;

[0084] Step S305: Obtain the deviation influence coefficient of any environmental factor on any radiation feature between any two abnormal records, calculate the average value to obtain the average influence coefficient of any environmental factor on any radiation feature; set the average influence coefficient of the jth environmental factor on the i2th radiation feature as X (j,i2) ; Randomly select an abnormal record and obtain the deviation amplitude of the jth environmental factor as η j The deviation from the i1th radiation feature in the hth radiation type is (f h ) i1 , according to the formula:

[0085]

[0086] Among them, i1, i2 and j are all positive integers, and i1∈(1,n), i2∈(1,n), j∈(1,m), n is the number of radiation features matched by the h-th radiation type, and m is the total number of environmental factors; the anomaly assessment value YP of the h-th radiation type in the selected anomaly record is calculated h ; Obtain the abnormal evaluation value of the hth radiation type in each abnormal record, and select the abnormal evaluation value with the smallest value as the abnormal evaluation threshold of the hth radiation type;

[0087] Example 2: Assume that there are two abnormal factors in an abnormal record and the deviation amplitudes η are 3% and 5% respectively, and there are two radiation types, one of which has two radiation features with deviation amplitudes f of 10% and 15% respectively, and the other radiation type has a radiation feature with a deviation amplitude f of 8%; at the same time, the average influence coefficients of the two abnormal factors on the three radiation features are 1%, 2% and 3% respectively, then in the abnormal record, the abnormality evaluation value YP1 of one radiation type is obtained as follows: (10% + 15%) + (3% × 1% + 3% × 2% + 5% × 1% + 5% × 2%) = 0.25 + 0.0003 + 0.0006 + 0.0005 + 0.001 = 0.2524, and the abnormality evaluation value YP2 of the other radiation type is obtained as follows: 8% + (3% × 3% + 5% × 3%) = 0.08 + 0.0009 + 0.0015 = 0.0824.

[0088] Step S400: generating a real-time detection record for the data collected in real time by the semiconductor detector, and performing an abnormality assessment on the real-time detection record; retrieving a corresponding abnormality assessment threshold for the real-time detection record, and determining whether to issue a risk warning for the real-time detection record;

[0089] Step S400 includes the following steps:

[0090] Step S401: Acquire detection data and environmental data collected in real time by the current semiconductor detector, obtain features and abnormal factors contained in the real-time detection record; compare the contained features with a radiation feature set in a radiation type database to obtain several radiation types of the real-time detection record;

[0091] Step S402: Obtain the deviation amplitude of any feature and the deviation amplitude of any abnormal factor in the various radiation types recorded in the real-time detection, extract the average influence coefficient of any abnormal factor on any feature, and calculate the abnormal evaluation value YP of the real-time detection record. now ;

[0092] Step S403: Obtain anomaly assessment thresholds of any radiation type in the real-time detection record, and sum them up to obtain an expected anomaly assessment threshold YP th , if YP now >YP th , then a risk warning is issued for the real-time detection record.

[0093] A data analysis system based on semiconductor detectors, the analysis system includes a historical detection analysis module, a radiation feature analysis module, an abnormality threshold division module, and a risk abnormality assessment module;

[0094] The historical detection analysis module is used to build a detection data management platform to receive the detection data and environmental data collected by the semiconductor detector during each detection process and generate corresponding detection records; conduct multi-dimensional evaluation of the detection data stored in any detection record and identify anomalies in any detection record;

[0095] The radiation feature analysis module is used to analyze the detection data corresponding to each dimension in any detection record and extract features from the detection data of each dimension; combine the various features extracted from the detection record and identify the radiation type present in the detection record;

[0096] The abnormality threshold classification module is used to perform differential analysis on any detection records that identify abnormalities and extract abnormal factors that affect various radiation types; based on the abnormal factors contained in different detection records, corresponding abnormality assessment thresholds are set for different radiation types;

[0097] The risk anomaly assessment module is used to generate a real-time detection record for the data collected by the semiconductor detector in real time, and perform an anomaly assessment on the real-time detection record; retrieve the corresponding anomaly assessment threshold for the real-time detection record, and determine whether to issue a risk warning for the real-time detection record.

[0098] Among them, the historical detection and analysis module includes a detection record collection unit and an abnormal record identification unit;

[0099] The detection record acquisition unit is used to build a detection data management platform to receive the detection data and environmental data collected by the semiconductor detector during each detection process and generate corresponding detection records; the abnormal record identification unit is used to carry out multi-dimensional evaluation of the detection data stored in any detection record and identify abnormalities in any detection record.

[0100] Among them, the radiation feature analysis module includes a radiation feature extraction unit and a radiation type identification unit;

[0101] The radiation feature extraction unit is used to analyze the detection data corresponding to each dimension in any detection record and extract features from the detection data of each dimension; the radiation type identification unit is used to combine the various features extracted from the detection record and identify the radiation type existing in the detection record.

[0102] Among them, the abnormal threshold division module includes an environmental factor extraction unit and a dynamic threshold adjustment unit;

[0103] The environmental factor extraction unit is used to perform differential analysis on any detection records that identify abnormalities and extract abnormal factors that affect various radiation types; the dynamic threshold adjustment unit is used to set corresponding abnormality assessment thresholds for different radiation types based on the abnormal factors contained in different detection records.

[0104] Among them, the risk anomaly assessment module includes a real-time risk assessment unit and an abnormal risk warning unit;

[0105] The real-time risk assessment unit is used to generate a real-time detection record for the data collected by the semiconductor detector in real time and perform an abnormality assessment on the real-time detection record; the abnormal risk warning unit is used to retrieve the corresponding abnormality assessment threshold for the real-time detection record and determine whether to issue a risk warning for the real-time detection record.

[0106] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims, not the foregoing description, and all variations within the meaning and range of equivalents of the claims are intended to be included therein. Any reference sign in a claim should not be construed as limiting the claim to which it relates.

Claims

1. A data analysis method based on a semiconductor detector, characterized in that: The analytical method comprises the following steps: Step S100: Build a detection data management platform to receive detection data and environmental data collected by each detection process of the semiconductor detector and generate corresponding detection records; conduct multi-dimensional evaluation of the detection data stored in any detection record and identify anomalies in any detection record; Step S200: Analyze the detection data corresponding to each dimension in any detection record, extract features from the detection data of each dimension; combine the features extracted from the detection record to identify the radiation type present in the detection record; Step S300: performing a difference analysis on any detected abnormality detection records to extract abnormality factors affecting various radiation types; setting corresponding abnormality assessment thresholds for different radiation types based on the abnormality factors contained in different detection records; Step S400: generating a real-time detection record for the data collected in real time by the semiconductor detector, and performing an abnormality assessment on the real-time detection record; retrieving a corresponding abnormality assessment threshold for the real-time detection record, and determining whether to issue a risk warning for the real-time detection record.

2. The data analysis method based on a semiconductor detector according to claim 1, characterized in that: The step S100 includes the following steps: Step S101: Several environmental sensors are pre-installed on the semiconductor detector. Whenever the semiconductor detector is used to detect the desired detection area, the data collected by the semiconductor detector is set as detection data, and the data collected by the several environmental sensors is set as environmental data. The detection data and environmental data are synchronously transmitted to the detection data management platform; Step S102: Summarize the probe data and environmental data received from the probe data management platform to generate a probe record; extract the probe data from any selected probe record, divide the probe data into several dimensions according to data type, preset corresponding evaluation rules for each dimension, and obtain the evaluation value of the probe data in any dimension; Step S103: Calculate the average value of the evaluation values under each dimension in the selected detection record to obtain the evaluation value of the selected detection record as P; preset an evaluation threshold P th , if P<P th , the selected detection record is set as an abnormal record, otherwise, the selected detection record is set as a normal record.

3. The data analysis method based on a semiconductor detector according to claim 2, characterized in that: The step S200 includes the following steps: Step S201: arbitrarily select a detection record, and arbitrarily select detection data under a dimension from the selected detection record to obtain the data range under the selected dimension, wherein the data range of the i-th dimension is set to (a i ,b i ); extract the preset evaluation rules of the i-th dimension and obtain the expected data range of the i-th dimension (a ex ,b ex ), if (a i ,b i )∈(a ex ,b ex ), the i-th dimension in the selected detection record is set as the normal dimension, otherwise, the i-th dimension is set as the abnormal dimension; Step S202: arbitrarily select an abnormal record. If the i-th dimension in the selected abnormal record is an abnormal dimension, extract features from the detection data of the i-th dimension; summarize the features of all abnormal dimensions in the selected abnormal record to obtain a feature set of the selected abnormal record; Step S203: Pre-establish a radiation type database, storing a plurality of radiation types and a plurality of radiation feature sets in the database, wherein any radiation type matches a radiation feature set; compare the feature set of the selected abnormal record with any radiation feature set; if any radiation feature in a radiation feature set is identical to a feature in the feature set, set the radiation type matched by the radiation feature set to be compared as a radiation type of the selected abnormal record, thereby obtaining a plurality of radiation types of the selected abnormal record; Step S204: Assume that any normal record contains several features, then compare the features with the radiation features in the radiation type database. If several radiation types are obtained after comparing the features of the normal record, then arbitrarily select one of the radiation types and extract all abnormal records containing the selected radiation type. Assume that the dimension where there is a radiation feature in the selected radiation type is the i-th dimension, and arbitrarily select an abnormal record from all the extracted abnormal records. The data range of the i-th dimension in the selected abnormal record is (a i ,b i ) and the expected data range is (a ex ,b ex ), the deviation amplitude of the i-th dimension in the selected abnormal records is f i =(b i -b ex ) / b ex ; Step S205: Obtain the deviation amplitude of the i-th dimension in the remaining abnormal records to obtain the abnormal deviation amplitude range of the i-th dimension ((f i ) min ,(f i ) max ); If there is a normal record containing the selected radiation type, the deviation amplitude of the i-th dimension is f1, and f1∈((f i ) min ,(f i ) max ), then the abnormal deviation range of the i-th dimension is corrected to (f1, (f i ) max ); Step S206: arbitrarily select a radiation type. If there is a normal record containing the selected radiation type, the selected radiation type and the corresponding abnormal deviation amplitude range are merged in the abnormal record containing the selected radiation type to generate a data set of the selected radiation type; if there is no normal record containing the selected radiation type, only the selected radiation type is generated in the abnormal record containing the selected radiation type to obtain a set of radiation types existing in any detection record.

4. The data analysis method based on a semiconductor detector according to claim 3, characterized in that: The step S300 includes the following steps: Step S301: The environmental data recorded in any detection record is divided into several categories of environmental data according to data type, and the corresponding environmental factors are extracted for each category of environmental data; an environmental factor is randomly selected, and the data range of the selected environmental factor in each normal record is obtained, and a union operation is performed to obtain the normal data range of the selected environmental factor; Step S302: Randomly select an abnormal record and obtain the data range (r1, r2) of the jth environmental factor, and set the normal data range of the jth environmental factor as R j ,like Then set the jth environmental factor as an abnormal factor to select abnormal records, and obtain the normal data range R j The minimum value (r j ) min and the maximum value (r j ) max , according to the formula: Among them, Max() is the maximum value function; the deviation amplitude η of the jth environmental factor in the selected abnormal record is calculated j ; Step S303: arbitrarily select two abnormal records containing the same radiation type, arbitrarily select a radiation feature from the same radiation type, and obtain the deviation amplitude of the two abnormal records in the abnormal dimension where the selected radiation feature is located. Assuming that the selected radiation feature is in the i-th abnormal dimension, the deviation amplitude difference Δf between the two abnormal records is obtained. i ; Step S304: Obtain each abnormal factor in the two abnormal records respectively, and obtain the deviation amplitude difference of the environmental factors corresponding to any abnormal factor between the two abnormal records, wherein the deviation amplitude difference of the jth environmental factor is set to Δη j , according to the formula: Where m is the total number of environmental factors; the deviation influence coefficient y of the jth environmental factor is calculated j ; Obtain the deviation amplitude difference between the two abnormal records of each radiation feature in the same radiation type, and obtain the deviation influence coefficient of the j-th environmental factor on each radiation feature; Step S305: Obtain the deviation influence coefficient of any environmental factor on any radiation feature between any two abnormal records, calculate the average value to obtain the average influence coefficient of any environmental factor on any radiation feature; set the average influence coefficient of the jth environmental factor on the i2th radiation feature as X (j,i2) ; Randomly select an abnormal record and obtain the deviation amplitude of the jth environmental factor as η j The deviation from the i1th radiation feature in the hth radiation type is (f h ) i1 , according to the formula: Among them, i1, i2 and j are all positive integers, and i1∈(1,n), i2∈(1,n), j∈(1,m), n is the number of radiation features matched by the h-th radiation type, and m is the total number of environmental factors; the anomaly assessment value YP of the h-th radiation type in the selected anomaly record is calculated h ; Obtain the abnormal evaluation value of the h-th radiation type in each abnormal record, and select the abnormal evaluation value with the smallest value as the abnormal evaluation threshold of the h-th radiation type.

5. The data analysis method based on semiconductor detector according to claim 4, characterized in that: The step S400 includes the following steps: Step S401: Acquire detection data and environmental data collected in real time by the current semiconductor detector, obtain features and abnormal factors contained in the real-time detection record; compare the contained features with a radiation feature set in a radiation type database to obtain several radiation types of the real-time detection record; Step S402: Obtain the deviation amplitude of any feature and the deviation amplitude of any abnormal factor in the various radiation types recorded in the real-time detection, extract the average influence coefficient of any abnormal factor on any feature, and calculate the abnormal evaluation value YP of the real-time detection record. now ; Step S403: Obtain anomaly assessment thresholds of any radiation type in the real-time detection record, and sum them up to obtain an expected anomaly assessment threshold YP th , if YP now >YP th , then a risk warning is issued for the real-time detection record.

6. A semiconductor detector-based data analysis system, configured to execute the semiconductor detector-based data analysis method according to any one of claims 1 to 5, characterized in that: The analysis system includes a historical detection analysis module, a radiation feature analysis module, an anomaly threshold division module, and a risk anomaly assessment module; The historical detection analysis module is used to build a detection data management platform to receive the detection data and environmental data collected by the semiconductor detector during each detection process and generate corresponding detection records; conduct multi-dimensional evaluation of the detection data stored in any detection record and identify anomalies in any detection record; The radiation feature analysis module is used to analyze the detection data corresponding to each dimension in any detection record, extract features from the detection data of each dimension; combine the features extracted from the detection record, and identify the radiation type present in the detection record; The abnormality threshold division module is used to perform differential analysis on any detection record that identifies an abnormality and extract abnormal factors that affect various radiation types; based on the abnormal factors contained in different detection records, corresponding abnormality assessment thresholds are set for different radiation types; The risk anomaly assessment module is used to generate a real-time detection record for the data collected by the semiconductor detector in real time, and perform an anomaly assessment on the real-time detection record; retrieve the corresponding anomaly assessment threshold for the real-time detection record, and determine whether to issue a risk warning for the real-time detection record.

7. The semiconductor detector-based data analysis system according to claim 6, characterized in that: The historical detection and analysis module includes a detection record collection unit and an abnormal record identification unit; The detection record acquisition unit is used to build a detection data management platform to receive the detection data and environmental data collected by the semiconductor detector during each detection process and generate corresponding detection records; the abnormal record identification unit is used to carry out multi-dimensional evaluation of the detection data stored in any detection record and identify abnormalities in any detection record.

8. The semiconductor detector-based data analysis system according to claim 6, characterized in that: The radiation feature analysis module includes a radiation feature extraction unit and a radiation type identification unit; The radiation feature extraction unit is used to analyze the detection data corresponding to each dimension in any detection record and extract features from the detection data of each dimension; the radiation type identification unit is used to combine the various features extracted from the detection record and identify the radiation type present in the detection record.

9. The semiconductor detector-based data analysis system according to claim 6, characterized in that: The abnormal threshold division module includes an environmental factor extraction unit and a dynamic threshold adjustment unit; The environmental factor extraction unit is used to perform differential analysis on any detection record that identifies an anomaly and extract the abnormal factors that affect various radiation types; the dynamic threshold adjustment unit is used to set corresponding abnormality assessment thresholds for different radiation types based on the abnormal factors contained in different detection records.

10. The semiconductor detector-based data analysis system according to claim 6, characterized in that: The risk anomaly assessment module includes a real-time risk assessment unit and an abnormal risk early warning unit; The real-time risk assessment unit is used to generate a real-time detection record for the data collected in real time by the semiconductor detector, and perform an abnormality assessment on the real-time detection record; The abnormal risk warning unit is used to retrieve the corresponding abnormal assessment threshold for the real-time detection record and determine whether to issue a risk warning for the real-time detection record.

Images