CN104951643B - A kind of energy spectrum processing method suitable for neutron activation elementary analysis - Google Patents
A kind of energy spectrum processing method suitable for neutron activation elementary analysis Download PDFInfo
- Publication number
- CN104951643B CN104951643B CN201410483577.2A CN201410483577A CN104951643B CN 104951643 B CN104951643 B CN 104951643B CN 201410483577 A CN201410483577 A CN 201410483577A CN 104951643 B CN104951643 B CN 104951643B
- Authority
- CN
- China
- Prior art keywords
- spectrum shape
- spectrum
- peak
- shape
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Abstract
The invention discloses a kind of energy spectrum processing method suitable for neutron activation elementary analysis, by the way of effective ratio pair is directly carried out with numerous spectrum shapes in spectrum shape knowledge base, the information such as composition and content of each component in the tested material corresponding to new measurement spectrum shape is obtained.By carrying out standardization processing, peak-seeking processing to spectrum shape, peak position is consistent or close property is searched out with the known element species and the spectrum shape of content with reference to effect and according to sequencing of similarity, select and be most compared with reference to three spectrum shapes of effect, obtain the information such as composition and content of each component in the tested material corresponding to new measurement spectrum shape.This method effectively prevent the drawbacks of being required for every time in original scaling method again to all having composed graphic data and the progress linear regression of corresponding element content or nonlinear regression, directly the spectrum shape of shape knowledge base is composed in supplement every time, knowledge base is simple to operate with surveying with benefit.
Description
Technical field
The present invention relates to a kind of data processing technique method of elementary analysis, and in particular to a kind of neutron activation elementary analysis
Data processing technique method.
Background technology
The industries such as building materials, coal, thermoelectricity, metallurgy, mine need to the elemental composition of loose unpacked material on belt conveyor and contained
Amount is analyzed, with Instructing manufacture.Multi-element analyser by the use of neutron activation Prompt gamma rays analytical technology is as a kind of main
On-line element analysis instrument, the continuous analysis to loose unpacked material on belt conveyor can be realized well.
The analysis principle of such current instrument, be all neutron exposure on material, occur with each element atomic nucleus in material
Reaction, different elements radiate different-energy and the Characteristic γ ray of intensity, by detecting that the energy of Characteristic γ ray recognizes member
Plain species, by detecting that the intensity of Characteristic γ ray draws constituent content.
And specific analysis method, then it is to use the peak area for the Characteristic γ ray of certain element to build to should be first
The mathematical modeling of cellulose content, then determines the parameter in mathematical modeling by way of surveying sample demarcation.If i.e. according to pair
The material of the dry known constituent content of group is measured, and known constituent content and the elemental characteristic peak area are substituted into mathematical modulo
Type, by linear regression or the processing of nonlinear regression, obtains the parameters in mathematical modeling.
But be the drawbacks of this method：Unknown element can not be analyzed, and for the element of unstructured mathematical modeling
It can not analyze；Simultaneously can not also material of the analytical element content outside the coverage of demarcation sample；And increase demarcation sample every time
After product, it is required for re-starting recurrence processing.
The content of the invention
For above-mentioned defect, the present invention proposes a kind of data processing technique of novel neutron activation elementary analysis
Method.This method has abandoned original scaling method, using directly with composing numerous spectrum shapes progress effective ratios pair in shape knowledge base
Mode, obtains the information such as composition and content of each component in the tested material corresponding to new measurement spectrum shape.
The execution basis of this method is to build the spectrum shape knowledge base of a standard and specification, that is, is required in spectrum shape knowledge base
Spectrum shape, with identical " energy-road location " corresponding relation, and each spectrum shape time of measuring will be identical.If right and wrong
The spectrum shape of specification, then need by composing fractal transform, such as by the processing such as conventional " peak drift correction " and " live time amendment "
Method, you can conversion is restored again into spectrum shape knowledge base after turning into specification spectrum shape.
Data processing technique method concrete scheme of the present invention is as follows：
The first step：When actual measurement spectrum shape needs analysis, actual measurement spectrum shape is standardized first, spectrum fractal transform will be surveyed
It is modified to each spectrum shape in spectrum shape knowledge base with consistent " energy-road location " corresponding relation and time of measuring；
The spectrum shape in spectrum shape knowledge base is required, with identical Zong Dao location numbers and " energy-road location " corresponding relation, and
The time of measuring of each spectrum shape will be identical.If the spectrum shape of non-standard, then need by composing fractal transform, such as " peak drift is repaiied
Just " and " live time amendment ", conversion is restored again into spectrum shape knowledge base after turning into specification spectrum shape.
Wherein, a variety of technical methods in order to realize spectrum fractal transform such as " peak drift correction " and " live time amendment " have a lot
Kind, well known to the professional person with certain experiences and grasp；Likewise, the peak-seeking algorithm of road location where determining peak position
And disposal skill, also it has been with well known to the professional person of certain experiences and grasping.
Spectrum shape knowledge base is characterised by：
Include two class files in spectrum shape knowledge base, one of which is the spectrum shape text for including corresponding element content information
Part, another is the index file for all spectrum shape files.
Compose in each spectrum shape files in shape knowledge base, include complete spectrum shape information and the contained various members of correspondence
Whole in the concrete content of element, spectrum shape knowledge base includes the spectrum shape of complete spectrum shape information and corresponding element content information
File, with identical Zong Dao location numbers and " energy-road location " corresponding relation, and each, the time of measuring of spectrum shape will be identical,
Specifically file format can be：If the Zong Dao locations number of the spectrum shape files is N, then file has N+103 group numerical value, preceding N groups
Numerical value for spectrum shape each road location from the correspondence energy counting of order successively from low to high, that is, compose shape from the 1st road to N roads
Count value；103 groups are followed successively by the content corresponding to each element arranged according to atomic number afterwards, right if the element is not contained
Answer position to insert " 0 ", if clearly contain the element but the constituent content it is unknown if inserted " -1 " in correspondence position；
For the index file of all spectrum shape files, then by the peak position confidence in whole spectrum shape files in spectrum shape knowledge base
Breath generation, concrete scheme is as follows：
1）It is numbered to composing shape in knowledge base；If there is n spectrum shape, since 1, to the last one is n；
2）Each Pu Xingzong roads location number is N, then index file includes N number of index data；Each index data is position
Number length is n binary number；
3）The producing method of i-th of index data is：If m-th of spectrum the i-th roads of Xing location is peak position, the m of the i-th data
Position is " 1 ", is otherwise " 0 "；Wherein, 1≤m≤n, 1≤i≤N.
Second step：Peak-seeking processing is carried out to the actual measurement spectrum shape after conversion, i.e., by peak-seeking algorithm, confirms the actual measurement after conversion
Compose quantity and each peak corresponding road location position at peak in shape；
3rd step：Similitude searching is carried out to each spectrum shape in spectrum shape knowledge base, " peak position is consistent " is carried out first and is found,
Whole spectrum shape files with following feature are searched out from spectrum shape knowledge base：The quantity at the peak in the spectrum shape files and every
Road location position corresponding to individual peak, quantity and each peak corresponding road location position with peak in the actual measurement spectrum shape after conversion is complete
Unanimously；
The particular content for carrying out the method that " peak position is consistent " is found is as follows：
1）Spectrum shape in spectrum shape knowledge base is numbered；If there is n spectrum shape, since 1, to the last one is n；
2）If Pu Xingzong roads location number is N roads, set up N number of " data "；Each data are that two that digits long is n enter
Number processed；
3）The producing method of i-th of data is：If m-th of spectrum the i-th roads of Xing location is peak position, the m positions of the i-th data
It is otherwise " 0 " for " 1 "；
4）Peak-seeking processing is carried out to the actual measurement to be analyzed spectrum shape, determine specific road location a1 where peak position, a2,
A3 ... ak, wherein k compose the total number at peak in shape for actual measurement；
5）By in N number of data, the data of peak position road location are constant in correspondence actual measurement spectrum shape, i.e., " data a1 ", " data a2 ",
" data a3 " ... " data ak " etc. retains constant；And remaining data then each all negate；
6）Value in the same position of newly-generated N group data is subjected to boolean " and " computing, one new two is obtained and enters
Number processed；And the value on which position of the binary number is " 1 ", then the spectrum shape for corresponding to the corresponding numbering in knowledge base seeks to seek
The result looked for.
Wherein it is determined that the peak-seeking algorithm and disposal skill of road location where peak position, have been the professional people with certain experiences
Scholar is known and grasps.
If the quantity for the spectrum shape for meeting " peak position is consistent " searching condition can be searched out in spectrum shape knowledge base more than three
It is individual, then directly carry out the 4th step；Otherwise system prompt " it is recommended that the spectrum shape is augmented into knowledge base ", and continuing in spectrum shape knowledge base
In do not meet " peak position is consistent " find condition spectrum shape in carry out " peak position containings " find, i.e., from compose shape knowledge base in searching
Provide whole spectrum shape files of following feature：Number of the quantity at the peak in the spectrum shape files than peak in the actual measurement spectrum shape after conversion
Amount is more, and there is corresponding embodiment the road location position at peak in the spectrum shape files in the actual measurement spectrum shape after conversion；
The particular content for carrying out the method that " peak position containing " is found is as follows：
1）Spectrum shape in spectrum shape knowledge base is numbered；If there is n spectrum shape, since 1, to the last one is n；
2）If Pu Xingzong roads location number is N roads, set up N number of " data "；Each data are that two that digits long is n enter
Number processed；
3）The producing method of i-th of data is：If m-th of spectrum the i-th roads of Xing location is peak position, the m positions of the i-th data
It is otherwise " 0 " for " 1 "；
4）Peak-seeking processing is carried out to the actual measurement to be analyzed spectrum shape, determine specific road location a1 where peak position, a2,
A3 ... ak, wherein k compose the total number at peak in shape for actual measurement；
5）By in N number of data, the k of peak position road location, will according to boolean " and " computing is carried out in correspondence actual measurement spectrum shape
" data a1 ", " data a2 ", " data a3 " ... " value in data ak " same position carries out boolean " and " computing, obtains
One new binary number；
6）Value on which position of the resulting binary number is " 1 ", then corresponds to corresponding numbering in knowledge base
Spectrum shape seeks to the result found.
Meet " peak position is consistent " searching condition if can be searched out in spectrum shape knowledge base and meet " peak position containing "
The total quantity of the spectrum shape of searching condition then directly carries out the 4th step more than three；Otherwise continue not meeting in spectrum shape knowledge base
" peak position is consistent ", which is found, carries out " peak position missing " searching in the spectrum shape of condition, i.e., searched out from spectrum shape knowledge base with such as
Whole spectrum shape files of lower feature：The quantity at the peak in the spectrum shape files is fewer than the quantity at peak in the actual measurement spectrum shape after conversion,
And there is corresponding embodiment in the actual measurement spectrum shape of the road location position at peak after the conversion in the spectrum shape files；
The particular content for carrying out the method that " peak position missing " is found is as follows：
1）Spectrum shape in spectrum shape knowledge base is numbered；If there is n spectrum shape, since 1, to the last one is n；
2）If Pu Xingzong roads location number is N roads, set up N number of " data "；Each data are that two that digits long is n enter
Number processed；
3）The producing method of i-th of data is：If m-th of spectrum the i-th roads of Xing location is peak position, the m positions of the i-th data
It is otherwise " 0 " for " 1 "；
4）Peak-seeking processing is carried out to the actual measurement to be analyzed spectrum shape, determine specific road location a1 where peak position, a2,
A3 ... ak, wherein k compose the total number at peak in shape for actual measurement；
5）By in N number of data, the value that correspondence actual measurement is composed into the same position of the data of peak position road location in shape carries out boolean
" or " computing obtains binary number A, i.e., " data a1 ", " data a2 ", " data a3 " ... " the value of data ak " same position
Carry out boolean " or " computing and obtain binary number A；By remaining N-k data after then each is all negated, the value of same position
Carry out boolean " and " computing and obtain binary number B；
6）Value in newly-generated binary number A and binary number B same position is subjected to boolean " and " computing, obtained
To a new binary number；And the value on which position of the binary number is " 1 ", then the corresponding numbering in knowledge base is corresponded to
Spectrum shape seek to find result.
If can be searched out in spectrum shape knowledge base and meet " peak position is consistent ", " peak position containing ", " peak position missing "
The total quantity of the spectrum shape of searching condition then proceeds the 4th step more than three；Otherwise termination analysis；
4th step：Similitude comparison is carried out to the similar spectrum shape found and sorted, i.e., will finally be found in the 3rd step
To spectrum shape files carry out similitude with the actual measurement spectrum shape after conversion respectively and compare, and enter according to the order of similarity from big to small
Row arrangement；
The particular content of the comparison sort method is as follows：
The spectrum shape being compared with actual measurement spectrum shape will be needed to carry out degree of approximation comparison operation with actual measurement spectrum shape one by one and obtain near
Like angle value P, then it is ranked up according to approximate angle value P size, the more big degrees of approximation for then illustrating two spectrum shapes of approximate angle value P are just
Higher, when two spectrum shapes are identical, approximate angle value P results are maximum 1；The formula for specifically calculating approximate angle value is as follows：
Wherein P is spectrum shape A and spectrum shape B Similarity value, and n is spectrum shape A and spectrum shape B Zong Dao locations number, aiIt is right in shape A to compose
Ying Dao locations i numerical value, biFor the numerical value of correspondence road location i in spectrum shape B.
5th step：Choose has the spectrum shapes of similarity highest three with the actual measurement spectrum shape after conversion, by with after conversion
Actual measurement spectrum shape once carries out " peak net area-constituent content linear regression " to each peak, analyzes corresponding to actual measurement spectrum shape
The content of each component in material；The actual measurement spectrum shape after three selected spectrum shapes and conversion is amounted in four spectrum shapes
Each peak carry out net area calculating, then successively to corresponding to the peak and the peak of phase people having a common goal location element carry out three spectrum shapes
Middle peak is washed one's face the linear regression of product and each corresponding constituent content, and the content and peak for establishing the element are washed one's face the pass of product
Be function, then by after conversion actual measurement spectrum shape to should road location peak wash one's face product substitute into relation function, calculate conversion
The content of the element corresponding to actual measurement spectrum shape afterwards；By carrying out such transport to the peak one by one in the actual measurement spectrum shape after conversion
Calculation is handled, the content of the various elements corresponding to actual measurement spectrum shape after being converted；
If having the spectrum shape that " peak position missing " is found in three selected spectrum shapes, deposit in actual measurement spectrum shape after the conversion
Can not all there be corresponding presence some peak positions in three selected spectrum shapes, i.e., can not analyze, then corresponding to the peak
The content of element will be without analysis.
Beneficial effect:
This method has abandoned original scaling method, using directly with composing numerous spectrum shapes progress effective ratios pair in shape knowledge base
Mode, obtain the information such as composition and content of each component in the tested material corresponding to new measurement spectrum shape.
This method effectively prevent in original scaling method to be required for again to all having composed graphic data and correspondingly every time
The drawbacks of constituent content carries out linear regression or nonlinear regression, directly the spectrum shape of shape knowledge base, knowledge are composed in supplement every time
Storehouse is simple to operate with surveying with benefit.
And the method for original demarcation is only capable of drawing every time the content of the element by demarcation processing, and for without
The content of the element of demarcation can not computing.The method of the present invention can be by being compared with the close spectrum shape in spectrum shape knowledge base
To obtaining the species and content of contained element, and spectrum shape knowledge base can augment at any time, and the spectrum shape after supplement is that may participate in analysis
Computing, without re-scaling every time.
Brief description of the drawings
Fig. 1 is a kind of flow chart of power spectrum knowledge base comparison method suitable for neutron activation elementary analysis of the present invention
Concrete application mode
It is described with reference to the accompanying drawings the particular content of the present invention.
The first step：When the spectrum shape measured in real time needs analysis, actual measurement spectrum shape is standardized first, spectrum will be surveyed
Fractal transform is modified to each spectrum shape in spectrum shape knowledge base with consistent " energy-road location " corresponding relation and time of measuring；
Include two class files in spectrum shape knowledge base, one of which is the spectrum shape text for including corresponding element content information
Part, another is the index file for all spectrum shape files.
Compose in each spectrum shape files in shape knowledge base, include complete spectrum shape information and the contained various members of correspondence
Whole in the concrete content of element, spectrum shape knowledge base includes the spectrum shape of complete spectrum shape information and corresponding element content information
File, with identical Zong Dao location numbers and " energy-road location " corresponding relation, and each, the time of measuring of spectrum shape will be identical,
Specifically file format can be：If the Zong Dao locations number of the spectrum shape files is N, then file has N+103 group numerical value, preceding N groups
Numerical value for spectrum shape each road location from the correspondence energy counting of order successively from low to high, that is, compose shape from the 1st road to N roads
Count value；103 groups are followed successively by the content corresponding to each element arranged according to atomic number afterwards, right if the element is not contained
Answer position to insert " 0 ", if clearly contain the element but the constituent content it is unknown if inserted " -1 " in correspondence position.
For the index file of all spectrum shape files, then by the peak position confidence in whole spectrum shape files in spectrum shape knowledge base
Breath generation, concrete scheme is as follows：
1）It is numbered to composing shape in knowledge base；If there is n spectrum shape, since 1, to the last one is n；
2）Each Pu Xingzong roads location number is N, then index file includes N number of index data；Each index data is position
Number length is n binary number；
3）The producing method of i-th of index data is：If m-th of spectrum the i-th roads of Xing location is peak position, the m of the i-th data
Position is " 1 ", is otherwise " 0 "；Wherein, 1≤m≤n, 1≤i≤N.
Wherein, a variety of technical methods in order to realize spectrum fractal transform such as " peak drift correction " and " live time amendment " have a lot
Kind, well known to the professional person with certain experiences and grasp；Likewise, the peak-seeking algorithm of road location where determining peak position
And disposal skill, also it has been with well known to the professional person of certain experiences and grasping.
Second step：Peak-seeking processing is carried out to the actual measurement spectrum shape after conversion, i.e., by peak-seeking algorithm, confirms the actual measurement after conversion
Compose quantity and each peak corresponding road location position at peak in shape；It is industry and peak-seeking algorithm known to the public has many kinds
There is the technology well known to experience personage in field；
3rd step：Similitude searching is carried out to each spectrum shape in spectrum shape knowledge base, " peak position is consistent " is carried out first and is found,
Whole spectrum shape files with following feature are searched out from spectrum shape knowledge base：The quantity at the peak in the spectrum shape files and every
Road location position corresponding to individual peak, quantity and each peak corresponding road location position with peak in the actual measurement spectrum shape after conversion is complete
Unanimously；
The particular content for the method that the progress " peak position is consistent " is found is as follows：
1）Spectrum shape in spectrum shape knowledge base is numbered；If there is n spectrum shape, since 1, to the last one is n；
2）If Pu Xingzong roads location number is N roads, set up N number of " data "；Each data are that two that digits long is n enter
Number processed；
3）The producing method of i-th of data is：If m-th of spectrum the i-th roads of Xing location is peak position, the m positions of the i-th data
It is otherwise " 0 " for " 1 "；
4）Peak-seeking processing is carried out to the actual measurement to be analyzed spectrum shape, determine specific road location a1 where peak position, a2,
A3 ... ak, wherein k compose the total number at peak in shape for actual measurement；
5）By in N number of data, the data of peak position road location are constant in correspondence actual measurement spectrum shape, i.e., " data a1 ", " data a2 ",
" data a3 " ... " data ak " etc. retains constant；And remaining data then each all negate；
6）Value in the same position of newly-generated N group data is subjected to boolean " and " computing, one new two is obtained and enters
Number processed；And the value on which position of the binary number is " 1 ", then the spectrum shape for corresponding to the corresponding numbering in knowledge base seeks to seek
The result looked for.
If the quantity that the spectrum shape for meeting " peak position is consistent " searching condition can be searched out in spectrum shape knowledge base is n1,
It is the spectrum shape total number in spectrum shape knowledge base that should have 0≤n1≤N, wherein N；Meet if can be searched out in spectrum shape knowledge base
" peak position is consistent " finds the quantity of the spectrum shape of condition more than three, i.e. n1 ＞ 3 then directly carry out the 4th step；Otherwise system is carried
Show " it is recommended that the spectrum shape is augmented into knowledge base ", and continue not meeting " peak position is consistent " searching condition in spectrum shape knowledge base
Compose and " peak position containing " searching is carried out in shape, i.e., whole spectrum shape files with following feature are searched out from spectrum shape knowledge base：
The quantity at the peak in the spectrum shape files is more than the quantity at peak in the actual measurement spectrum shape after conversion, and in the actual measurement spectrum shape after conversion
There is corresponding embodiment the road location position at peak in the spectrum shape files；
The particular content for the method that the progress " peak position containing " is found is as follows：
1）Spectrum shape in spectrum shape knowledge base is numbered；If there is n spectrum shape, since 1, to the last one is n；
2）If Pu Xingzong roads location number is N roads, set up N number of " data "；Each data are that two that digits long is n enter
Number processed；
3）The producing method of i-th of data is：If m-th of spectrum the i-th roads of Xing location is peak position, the m positions of the i-th data
It is otherwise " 0 " for " 1 "；
4）Peak-seeking processing is carried out to the actual measurement to be analyzed spectrum shape, determine specific road location a1 where peak position, a2,
A3 ... ak, wherein k compose the total number at peak in shape for actual measurement；
5）By in N number of data, the k of peak position road location, will according to boolean " and " computing is carried out in correspondence actual measurement spectrum shape
" data a1 ", " data a2 ", " data a3 " ... " value in data ak " same position carries out boolean " and " computing, obtains
One new binary number；
6）Value on which position of the resulting binary number is " 1 ", then corresponds to corresponding numbering in knowledge base
Spectrum shape seeks to the result found.
If the quantity that the spectrum shape for meeting " peak position containing " searching condition can be searched out in spectrum shape knowledge base is n2,
There should be 0≤n2≤N-n1；Meet " peak position is consistent " searching condition if can be searched out in spectrum shape knowledge base and meet " peak position
Put containing " find condition spectrum shape total quantity more than three, i.e. n1+n2 ＞ 3, then directly progress the 4th step；Otherwise continue
" peak position missing " is carried out in the spectrum shape for not meeting " peak position is consistent " searching condition in spectrum shape knowledge base to find, i.e., know from spectrum shape
Know and whole spectrum shape files with following feature are searched out in storehouse：The quantity at the peak in the spectrum shape files is composed than the actual measurement after conversion
The quantity at peak will be lacked in shape, and in the spectrum shape files peak road location position after the conversion actual measurement spectrum shape in have corresponding body
It is existing；
The particular content for the method that the progress " peak position missing " is found is as follows：
1）Spectrum shape in spectrum shape knowledge base is numbered；If there is n spectrum shape, since 1, to the last one is n；
2）If Pu Xingzong roads location number is N roads, set up N number of " data "；Each data are that two that digits long is n enter
Number processed；
3）The producing method of i-th of data is：If m-th of spectrum the i-th roads of Xing location is peak position, the m positions of the i-th data
It is otherwise " 0 " for " 1 "；
4）Peak-seeking processing is carried out to the actual measurement to be analyzed spectrum shape, determine specific road location a1 where peak position, a2,
A3 ... ak, wherein k compose the total number at peak in shape for actual measurement；
5）By in N number of data, the value that correspondence actual measurement is composed into the same position of the data of peak position road location in shape carries out boolean
" or " computing obtains binary number A, i.e., " data a1 ", " data a2 ", " data a3 " ... " the value of data ak " same position
Carry out boolean " or " computing and obtain binary number A；By remaining N-k data after then each is all negated, the value of same position
Carry out boolean " and " computing and obtain binary number B；
6）Value in newly-generated binary number A and binary number B same position is subjected to boolean " and " computing, obtained
To a new binary number；And the value on which position of the binary number is " 1 ", then the corresponding numbering in knowledge base is corresponded to
Spectrum shape seek to find result.
If the quantity that the spectrum shape for meeting " peak position missing " searching condition can be searched out in spectrum shape knowledge base is n3,
There should be 0≤n3≤N-n1-n2；If can spectrum shape knowledge base in search out meet " peak position is consistent ", " peak position containing ",
" peak position missing " finds the total quantity of the spectrum shape of condition more than three, i.e. n1+n2+n3 ＞ 3 then proceed the 4th step；It is no
Then termination analysis；
4th step：Similitude comparison is carried out to the similar spectrum shape found and sorted, i.e., will finally be found in the 3rd step
To spectrum shape files carry out similitude with the actual measurement spectrum shape after conversion respectively and compare, and enter according to the order of similarity from big to small
Row arrangement；
The particular content of the comparison sort method is as follows：
The spectrum shape being compared with actual measurement spectrum shape will be needed to carry out degree of approximation comparison operation with actual measurement spectrum shape one by one and obtain near
Like angle value P, then it is ranked up according to approximate angle value P size, the more big degrees of approximation for then illustrating two spectrum shapes of approximate angle value P are just
Higher, when two spectrum shapes are identical, approximate angle value P results are maximum 1；The formula for specifically calculating approximate angle value is as follows：
Wherein P is spectrum shape A and spectrum shape B Similarity value, and n is spectrum shape A and spectrum shape B Zong Dao locations number, aiIt is right in shape A to compose
Ying Dao locations i numerical value, biFor the numerical value of correspondence road location i in spectrum shape B.
And spectrum shape conventional at present is typically all the total approximate angle value for 1024 roads, then 1024 roads of specific correspondence in location
Calculation formula be：
Similar, this method can be also used near between the spectrum shape of 512 roads, 2048 roads, 4096 roads etc. Zong Dao locations number
Compare like degree, but the degree of approximation being not merely only limited between the spectrum shape of above-described Zong Dao locations number compares.
5th step：Choose has the spectrum shapes of similarity highest three with the actual measurement spectrum shape after conversion, by with after conversion
Actual measurement spectrum shape once carries out " peak net area-constituent content linear regression " to each peak, analyzes corresponding to actual measurement spectrum shape
The content of each component in material；The actual measurement spectrum shape after three selected spectrum shapes and conversion is amounted in four spectrum shapes
Each peak carry out net area calculating, then successively to corresponding to the peak and the peak of phase people having a common goal location element carry out three spectrum shapes
Middle peak is washed one's face the linear regression of product and each corresponding constituent content, and the content and peak for establishing the element are washed one's face the pass of product
Be function, then by after conversion actual measurement spectrum shape to should road location peak wash one's face product substitute into relation function, calculate conversion
The content of the element corresponding to actual measurement spectrum shape afterwards；By carrying out such transport to the peak one by one in the actual measurement spectrum shape after conversion
Calculation is handled, the content of the various elements corresponding to actual measurement spectrum shape after being converted；
If having the spectrum shape that " peak position missing " is found in three selected spectrum shapes, deposit in actual measurement spectrum shape after the conversion
Can not all there be corresponding presence some peak positions in three selected spectrum shapes, i.e., can not analyze, then corresponding to the peak
The content of element will be without analysis.
Concrete application example：
Because in actual application, the spectrum figurate number amount that spectrum shape knowledge place is included is very huge, and each
The road location number of spectrum shape is generally 1024,2048 or even 4096 roads, for sake of convenience, therefore to several crucial processing in this patent
Method is illustrated, and all carries out hypothesis condition, but can't influence in industry experienced person to the complete of this method
Foliation solution.
First, the illustration of shape knowledge base is composed on building：
Because in actual application, the actual measurement Pu Xingzong roads location number of usual neutron activation multielement analysis is 1024
Road or 2048 roads, the road in the roads of Ye You 512,4096 roads or other quantity, to simplify the narration to application example, to spectrum shape files
And corresponding element is assumed as follows：
Assuming that the roads of Pu Xingzong roads location Shuo Wei 10, and in the absence of 103 kinds of elements, only exist atomic number be respectively 1,2,
3 element species, and have the position correspondence road location 2 of the characteristic peak for the element that atomic number is 1, atomic number is 2 element
Characteristic peak position correspondence road location 5, atomic number for 3 element characteristic peak position correspondence road location 8；
Assuming that the content of the data and corresponding each element on existing spectrum shape five, each of which road location see the table below：
Marked in table "" it is to represent that the content of the element is unknown.
Accordingly, the peak position in five spectrum shapes see the table below：
Then each spectrum shape files in spectrum shape knowledge base should have 13 groups of numerical value, specifically see the table below：
Then the index file in spectrum shape knowledge base has 10 index datas, and each index data is that two that digit is 5 enter
Number processed, specifically see the table below：
2nd, the illustration that " peak position is consistent " is found is carried out：
Because the Pu Xingzong roads location number in practical application is usually 1024 roads or 2048 roads, also there are 4096 roads, in order to
Narration is convenient, sets Pu Xingzong roads location number as 7 roads, and composes the spectrum shape number in shape knowledge base for 5, and specific setting is as follows：
Road location number is where composing the peak position of shape 1：1st, 3,5,7, i.e. peak position element is：{1、3、5、7}；
Road location number is where composing the peak position of shape 2：2nd, 4,6, i.e. peak position element is：{2、4、6}；
Road location number is where composing the peak position of shape 3：3rd, 5, i.e. peak position element is：{3、5}；
Road location number is where composing the peak position of shape 4：2nd, 5, i.e. peak position element is：{2、5}；
Road location number is where composing the peak position of shape 5：1st, 3,5, i.e. peak position element is：{1、3、5}；
It can see the table below in detail：
The corresponding data for forming 7 groups of binary numbers, specifically see the table below：
As can be seen from the above table：The content of data 1 is " 10001 ", then it is to have peak position to have embodied spectrum shape 1 and spectrum shape 5
Zhi roads location 1；Similar, the content of data 6 is " 01000 ", then having embodied spectrum shape 2 has peak position in road location 6.
Then start to carry out " peak position comparison processing " with newly spectrum shape：
If road location where having the peak position of new spectrum shape 8 is 1,3,5, can by " data 1 ", " data 3 ", " data 5 " are constant,
" data 2 ", " data 4 ", " data 6 ", " data 7 " carry out every negate；
Then the same position to all data carries out boolean " and " computing, obtains a new binary number, and is somebody's turn to do
Which position of binary number is " 1 ", then has which spectrum shape in spectrum shape knowledge base consistent with new spectrum shape.
Concrete operation see the table below：
It can be seen from the table, meet " peak position is consistent " searching condition with the peak position of new spectrum shape 8 is composed in shape knowledge base
" spectrum shape 5 ".
3rd, the illustration that " peak position containing " is found is carried out：
Because the Pu Xingzong roads location number in practical application is usually 1024 roads or 2048 roads, also there are 4096 roads, in order to
Narration is convenient, sets Pu Xingzong roads location number as 7 roads, and composes the spectrum shape number in shape knowledge base for 5, and specific setting is as follows：
Road location number is where composing the peak position of shape 1：1st, 3,5,7, i.e. peak position element is：{1、3、5、7}；
Road location number is where composing the peak position of shape 2：2nd, 4,6, i.e. peak position element is：{2、4、6}；
Road location number is where composing the peak position of shape 3：3rd, 5, i.e. peak position element is：{3、5}；
Road location number is where composing the peak position of shape 4：2nd, 5, i.e. peak position element is：{2、5}；
Road location number is where composing the peak position of shape 5：1st, 3,5, i.e. peak position element is：{1、3、5}；
It can see the table below in detail：
The corresponding data for forming 7 groups of binary numbers, specifically see the table below：
As can be seen from the above table：The content of data 1 is " 10001 ", then it is to have peak position to have embodied spectrum shape 1 and spectrum shape 5
Zhi roads location 1；Similar, the content of data 6 is " 01000 ", then having embodied spectrum shape 2 has peak position in road location 6.
Then start to carry out " peak position comparison processing " with newly spectrum shape：
If road location where having the peak position of new spectrum shape 8 is 1,3,5, can be by " data 1 ", " data 3 ", " phase of data 5 "
Boolean " and " computing is carried out with position, a new binary number is obtained, and which position of the binary number is " 1 ", then
There is which spectrum shape in spectrum shape knowledge base consistent with new spectrum shape.
Concrete operation see the table below：
It can be seen from the table, meet " peak position containing " searching condition with the peak position of new spectrum shape 8 is composed in shape knowledge base
" spectrum shape 1 " and " spectrum shape 5 ".
It is special to be emphasized that in the method, it can will meet the spectrum shape searching that " peak position is consistent " finds condition
Out, thus this method need spectrum shape knowledge base in do not meet " peak position is consistent " find condition spectrum shape in carry out.
4th, the illustration that " peak position missing " is found is carried out：
Because the Pu Xingzong roads location number in practical application is usually 1024 roads or 2048 roads, also there are 4096 roads, in order to
Narration is convenient, sets Pu Xingzong roads location number as 7 roads, and composes the spectrum shape number in shape knowledge base for 5, and specific setting is as follows：
Road location number is where composing the peak position of shape 1：1st, 3,5,7, i.e. peak position element is：{1、3、5、7}；
Road location number is where composing the peak position of shape 2：2nd, 4,6, i.e. peak position element is：{2、4、6}；
Road location number is where composing the peak position of shape 3：3rd, 5, i.e. peak position element is：{3、5}；
Road location number is where composing the peak position of shape 4：2nd, 5, i.e. peak position element is：{2、5}；
Road location number is where composing the peak position of shape 5：1st, 3,5, i.e. peak position element is：{1、3、5}；
It can see the table below in detail：
The corresponding data for forming 7 groups of binary numbers, specifically see the table below：
As can be seen from the above table：The content of data 1 is " 10001 ", then it is to have peak position to have embodied spectrum shape 1 and spectrum shape 5
Zhi roads location 1；Similar, the content of data 6 is " 01000 ", then having embodied spectrum shape 2 has peak position in road location 6.
Then start to carry out " peak position comparison processing " with newly spectrum shape：
If road location where having the peak position of new spectrum shape 8 is 1,3,5, can be by " data 1 ", " data 3 ", " phase of data 5 "
Boolean " or " computing is carried out with the value on position, new data A is obtained for " 10111 "；Concrete operation see the table below：
Again will " data 2 ", " data 4 ", " data 6 ", " data 7 " carry out every negate after, then to the value in same position
Boolean " and " computing is carried out, new data B is obtained for " 00101 "；Concrete operation see the table below：
Then boolean " and " computing is carried out to the value in new data A and new data B same position again, obtains one newly
Binary number, and which position of the binary number is " 1 ", then has in spectrum shape knowledge base which spectrum shape with newly composing shape phase
Unanimously.
Concrete operation see the table below：
It can be seen from the table, meet " peak position missing " searching condition with the peak position of new spectrum shape 8 is composed in shape knowledge base
" spectrum shape 3 " and " spectrum shape 5 ".
It is special to be emphasized that in the method, it can will meet the spectrum shape searching that " peak position is consistent " finds condition
Out, thus this method need spectrum shape knowledge base in do not meet " peak position is consistent " find condition spectrum shape in carry out.
5th, similitude comparison and the illustration sorted are carried out on spectrum shape：
Because the Pu Xingzong roads location number in practical application is usually 1024 roads or 2048 roads, also there are 4096 roads, in order to
Narration is convenient, and this concrete application example sets Pu Xingzong roads location number as 9 roads, there is spectrum shape A, spectrum shape B, spectrum shape C, spectrum shape D respectively
The degree of approximation is carried out with spectrum shape O to be compared.That is spectrum shape O is actual measurement spectrum shape, and it is from spectrum shape knowledge to compose shape A, spectrum shape B, spectrum shape C, spectrum shape D
The four spectrum shapes chosen in storehouse are, it is necessary to spectrum shape A, spectrum shape B, spectrum shape C, approximate journey of this four spectrum shapes of spectrum shape D relative to spectrum shape O
Degree is ranked up.
Setting spectrum shape O, spectrum shape A, spectrum shape B, spectrum shape C, the spectrum shape D corresponding numerical value in each road location see the table below：
Then have：
Shape A is composed with composing shape O approximate angle value POAFor：
；
Shape B is composed with composing shape O approximate angle value POBFor：
；
Shape C is composed with composing shape O approximate angle value POCFor：
；
Shape D is composed with composing shape O approximate angle value PODFor：
；
Have：
POA=0.9927；
POB=0.9869；
POC=0.8744；
POD=1.0000；
It can be seen that：POD＞ POA＞ POB＞ POC。
Then it is ordered as according to approximate angle value is descending：Four spectrum shapes are descending relative to spectrum shape O degree of approximation
Put in order is spectrum shape D, spectrum shape A, spectrum shape B, spectrum shape C.Wherein spectrum shape D is identical with spectrum shape O.
Claims (6)
1. a kind of energy spectrum processing method suitable for neutron activation elementary analysis, it is characterized in that：
The first step：When actual measurement spectrum shape needs analysis, actual measurement spectrum shape is standardized first, spectrum fractal transform amendment will be surveyed
For there is consistent Zong Dao locations number, " energy-road location " corresponding relation and time of measuring with each spectrum shape in spectrum shape knowledge base；
Second step：Peak-seeking processing is carried out to the actual measurement spectrum shape after conversion, i.e., by peak-seeking algorithm, confirms the actual measurement spectrum shape after conversion
The quantity at middle peak and each peak corresponding road location position；
3rd step：To spectrum shape knowledge base in each spectrum shape carry out similitude searching, first carry out " peak position is consistent " find, i.e., from
Whole spectrum shape files with following feature are searched out in spectrum shape knowledge base：The quantity and each peak at the peak in the spectrum shape files
Corresponding road location position, quantity and each peak corresponding road location position complete one with peak in the actual measurement spectrum shape after conversion
Cause；
If the quantity for the spectrum shape for meeting " peak position is consistent " searching condition can be searched out in spectrum shape knowledge base more than three,
Directly carry out the 4th step；Otherwise system prompt " it is recommended that the spectrum shape is augmented into knowledge base ", and continuing in spectrum shape knowledge base not
" peak position containing " is carried out in the spectrum shape for meeting " peak position is consistent " searching condition to find, i.e., find and provide from spectrum shape knowledge base
There are whole spectrum shape files of following feature：The quantity at the peak in the spectrum shape files will than the quantity at peak in the actual measurement spectrum shape after conversion
It is many, and there is corresponding embodiment the road location position at peak in the spectrum shape files in actual measurement spectrum shape after converting；
Found if can search out to meet " peak position is consistent " searching condition and meet " peak position containing " in spectrum shape knowledge base
The total quantity of the spectrum shape of condition then directly carries out the 4th step more than three；Otherwise continue not meeting " peak position in spectrum shape knowledge base
Put consistent " find and carry out " peak position missing " in the spectrum shape of condition and find, i.e., searched out from spectrum shape knowledge base with spy as follows
The whole spectrum shape files levied：The quantity at the peak in the spectrum shape files is fewer than the quantity at peak in the actual measurement spectrum shape after conversion, and
There is corresponding embodiment in the actual measurement spectrum shape of the road location position at peak after the conversion in the spectrum shape files；
Meet " peak position is consistent ", " peak position containing ", " peak position missing " searching if can be searched out in spectrum shape knowledge base
The total quantity of the spectrum shape of condition then proceeds the 4th step more than three；Otherwise termination analysis；
4th step：Similitude comparison is carried out to the similar spectrum shape that finds and sort, i.e., will finally be searched out in the 3rd step
Spectrum shape files carry out similitude with the actual measurement spectrum shape after conversion respectively and compared, and are arranged according to the order of similarity from big to small
Row；
5th step：Choose has the spectrum shapes of similarity highest three with the actual measurement spectrum shape after conversion, by with the actual measurement after conversion
Spectrum shape carries out " peak net area-constituent content linear regression " to each peak successively, analyzes the material corresponding to actual measurement spectrum shape
In each component content；The actual measurement spectrum shape after three selected spectrum shapes and conversion is amounted to each in four spectrum shapes
Individual peak carries out net area calculating, then carries out peak in three spectrum shapes to the element corresponding to the peak and the peak of phase people having a common goal location successively
Wash one's face the linear regression of product and each corresponding constituent content, the content and peak for establishing the element are washed one's face the relation letter of product
Number, then by after conversion actual measurement spectrum shape to should road location peak wash one's face product substitute into relation function, calculate conversion after
The content of the element corresponding to actual measurement spectrum shape；By being carried out to the peak one by one in the actual measurement spectrum shape after conversion at such computing
Reason, the content of the various elements corresponding to actual measurement spectrum shape after being converted；
If having the spectrum shape that " peak position missing " is found in three selected spectrum shapes, deposit some in actual measurement spectrum shape after the conversion
Can not all there be corresponding presence peak position in three selected spectrum shapes, i.e., can not analyze, then the element corresponding to the peak
Content will without analysis.
2. a kind of energy spectrum processing method suitable for neutron activation elementary analysis according to claim 1, it is characterised in that：
Include two class files in described spectrum shape knowledge base, one of which is the spectrum shape text for including corresponding element content information
Part, another is the index file for all spectrum shape files；
Whole in spectrum shape knowledge base includes the spectrum shape files of corresponding element content information, with identical Zong Dao location numbers and
" energy-road location " corresponding relation, and each the time of measuring of spectrum shape will be identical；
And specific file format is：If the Pu Xingzong roads location number is N, then file has N+103 group numerical value, preceding N groups
Numerical value for spectrum shape each road location from the correspondence energy counting of order successively from low to high, that is, compose shape from the 1st road to N roads
Count value；103 groups are followed successively by the content corresponding to each element arranged according to atomic number afterwards, right if the element is not contained
Answer position to insert " 0 ", if clearly contain the element but the constituent content it is unknown if inserted " -1 " in correspondence position；
For the index file of all spectrum shape files, given birth to by the peak position information in whole spectrum shape files in spectrum shape knowledge base
Into concrete scheme is as follows：
1）It is numbered to composing shape in knowledge base；If there is n spectrum shape, since 1, to the last one is n；
2）Each Pu Xingzong roads location number is N, then index file includes N number of index data；Each index data is that digit is long
Degree is n binary number；
3）The producing method of i-th of index data is：If m-th of spectrum the i-th roads of Xing location is peak position, the m positions of the i-th data
It is otherwise " 0 " for " 1 "；Wherein, 1≤m≤n, 1≤i≤N.
3. a kind of energy spectrum processing method suitable for neutron activation elementary analysis according to claim 1, it is characterised in that：
Described " peak position is consistent " is found, and the detailed process for finding the method for spectrum shape is as follows：
1）Spectrum shape in spectrum shape knowledge base is numbered, if there is n spectrum shape, since 1, to the last one is n；
2）If Pu Xingzong roads location number is N roads, set up N number of " data "；Each data are the binary systems that digits long is n
Number；
3）The producing method of i-th of data is：If m-th of spectrum the i-th roads of Xing location is peak position, the m positions of the i-th data are
" 1 ", is otherwise " 0 "；
4）Peak-seeking processing is carried out to the actual measurement to be analyzed spectrum shape, determine specific road location a1 where peak position, a2, a3 ... ak,
Wherein k composes the total number at peak in shape for actual measurement；
5）By in N number of data, the data of peak position road location are constant in correspondence actual measurement spectrum shape, i.e. " data a1 ", " data a2 ", " number
According to a3 " ... " data ak " etc. retain it is constant；And remaining data then each all negate；
6）Value in the same position of newly-generated N group data is subjected to boolean " and " computing, a new binary system is obtained
Number；And the value on which position of the binary number is " 1 ", then the spectrum shape for corresponding to the corresponding numbering in knowledge base seeks to find
Result.
4. a kind of energy spectrum processing method suitable for neutron activation elementary analysis according to claim 1, it is characterised in that：
Described " peak position containing " is found, and the detailed process for finding the method for spectrum shape is as follows：
1）Spectrum shape in spectrum shape knowledge base is numbered；If there is n spectrum shape, since 1, to the last one is n；
2）If Pu Xingzong roads location number is N roads, set up N number of " data "；Each data are the binary systems that digits long is n
Number；
3）The producing method of i-th of data is：If m-th of spectrum the i-th roads of Xing location is peak position, the m positions of the i-th data are
" 1 ", is otherwise " 0 "；
4）Peak-seeking processing is carried out to the actual measurement to be analyzed spectrum shape, determine specific road location a1 where peak position, a2, a3 ... ak,
Wherein k composes the total number at peak in shape for actual measurement；
5）By in N number of data, the k of peak position road location, will " number according to boolean " and " computing is carried out in correspondence actual measurement spectrum shape
According to a1 ", " data a2 ", " data a3 " ... " value in data ak " same position carries out boolean " and " computing, obtains one
Individual new binary number；
6）Value on which position of the resulting binary number is " 1 ", then corresponds to the spectrum shape of the corresponding numbering in knowledge base
Seek to the result found.
5. a kind of energy spectrum processing method suitable for neutron activation elementary analysis according to claim 1, it is characterised in that：
Described " peak position missing " is found, and the detailed process for finding the method for spectrum shape is as follows：
1）Spectrum shape in spectrum shape knowledge base is numbered；If there is n spectrum shape, since 1, to the last one is n；
2）If Pu Xingzong roads location number is N roads, set up N number of " data "；Each data are the binary systems that digits long is n
Number；
3）The producing method of i-th of data is：If m-th of spectrum the i-th roads of Xing location is peak position, the m positions of the i-th data are
" 1 ", is otherwise " 0 "；
4）Peak-seeking processing is carried out to the actual measurement to be analyzed spectrum shape, determine specific road location a1 where peak position, a2, a3 ... ak,
Wherein k composes the total number at peak in shape for actual measurement；
5）By in N number of data, the value that correspondence actual measurement is composed into the same position of the data of peak position road location in shape carries out boolean " or "
Computing obtains binary number A, i.e., " data a1 ", " data a2 ", " data a3 " ... " value of data ak " same position is carried out
Boolean " or " computing obtains binary number A；By remaining N-k data after then each is all negated, the value of same position is carried out
Boolean " and " computing obtains binary number B；
6）Value in newly-generated binary number A and binary number B same position is subjected to boolean " and " computing, one is obtained
Individual new binary number；And the value on which position of the binary number is " 1 ", then the spectrum of the corresponding numbering in knowledge base is corresponded to
Shape seeks to the result found.
6. a kind of energy spectrum processing method suitable for neutron activation elementary analysis according to claim 1, it is characterised in that：
Described to carry out similitude comparison to the similar spectrum shape that finds and sort, method is：
The spectrum shape being compared with actual measurement spectrum shape will be needed to carry out degree of approximation comparison operation with actual measurement spectrum shape one by one and obtain the degree of approximation
Value P, is then ranked up according to approximate angle value P size, and the more big degrees of approximation for then illustrating two spectrum shapes of approximate angle value P are more
Height, when two spectrum shapes are identical, approximate angle value P results are maximum 1；The formula for specifically calculating approximate angle value is as follows：
Wherein P is spectrum shape A and spectrum shape B Similarity value, and n is spectrum shape A and spectrum shape B Zong Dao locations number, aiFor correspondence road in spectrum shape A
Location i numerical value, biFor the numerical value of correspondence road location i in spectrum shape B.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410483577.2A CN104951643B (en) | 2014-03-27 | 2014-09-20 | A kind of energy spectrum processing method suitable for neutron activation elementary analysis |
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410117363 | 2014-03-27 | ||
| CN2014101173633 | 2014-03-27 | ||
| CN201410483577.2A CN104951643B (en) | 2014-03-27 | 2014-09-20 | A kind of energy spectrum processing method suitable for neutron activation elementary analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104951643A CN104951643A (en) | 2015-09-30 |
| CN104951643B true CN104951643B (en) | 2017-09-26 |
Family
ID=54166153
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410483577.2A Active CN104951643B (en) | 2014-03-27 | 2014-09-20 | A kind of energy spectrum processing method suitable for neutron activation elementary analysis |
| CN201410592213.8A Active CN104951490B (en) | 2014-03-27 | 2014-10-30 | A kind of power spectrum knowledge base application process suitable for elementary analysis |
Family Applications After (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410592213.8A Active CN104951490B (en) | 2014-03-27 | 2014-10-30 | A kind of power spectrum knowledge base application process suitable for elementary analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN104951643B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107290355A (en) * | 2016-03-31 | 2017-10-24 | 丹东东方测控技术股份有限公司 | The online neutron activation alumina silica ratio analytical equipment of transmission-type bauxite and analysis method |
| CN107490592B (en) * | 2017-08-18 | 2020-08-28 | 中国工程物理研究院核物理与化学研究所 | On-line sample activating and analyzing device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1588020A (en) * | 2004-10-15 | 2005-03-02 | 南京大陆中电科技股份有限公司 | Coal element analytic method and on-line detecting equipment based on spectral library least square method |
| CN1632544A (en) * | 2004-12-30 | 2005-06-29 | 吉林省科仑辐射技术开发有限公司 | Neutron cement multi-element analyzer |
| KR101207869B1 (en) * | 2011-12-28 | 2012-12-04 | 부산대학교 산학협력단 | Prediction method of delamination initiation and growth of adhesives by fem |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6157034A (en) * | 1998-07-02 | 2000-12-05 | Gamma-Metrics | Flexible multi-purpose modular assembly for a family of PGNAA bulk material analyzers |
| US7006919B2 (en) * | 2003-06-02 | 2006-02-28 | Energy Technologies, Inc. | Real time continuous elemental measurement of bulk material |
| CN1834632B (en) * | 2006-04-17 | 2011-03-16 | 南京大陆中电科技股份有限公司 | Concrete material content on-line detection system based on pulse fast heating neutron instantaneous Gamna radiation analysis technology |
-
2014
- 2014-09-20 CN CN201410483577.2A patent/CN104951643B/en active Active
- 2014-10-30 CN CN201410592213.8A patent/CN104951490B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1588020A (en) * | 2004-10-15 | 2005-03-02 | 南京大陆中电科技股份有限公司 | Coal element analytic method and on-line detecting equipment based on spectral library least square method |
| CN1632544A (en) * | 2004-12-30 | 2005-06-29 | 吉林省科仑辐射技术开发有限公司 | Neutron cement multi-element analyzer |
| KR101207869B1 (en) * | 2011-12-28 | 2012-12-04 | 부산대학교 산학협력단 | Prediction method of delamination initiation and growth of adhesives by fem |
Non-Patent Citations (2)
| Title |
|---|
| X荧光在线多元素分析仪的研究与应用;陈树军 等;《世界有色金属》;20131031(第10期);第76-77页 * |
| 数字多道技术在X荧光多元素分析仪中的应用;于海明 等;《测量仪器》;20130228(第1期);第55-57页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104951643A (en) | 2015-09-30 |
| CN104951490B (en) | 2018-01-30 |
| CN104951490A (en) | 2015-09-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Butter et al. | Deep-learned top tagging with a Lorentz layer | |
| Fraser et al. | Jet charge and machine learning | |
| Choma et al. | Graph neural networks for icecube signal classification | |
| Komiske et al. | An operational definition of quark and gluon jets | |
| Kazienko et al. | Hybrid and ensemble methods in machine learning J. UCS special issue | |
| Endah et al. | Xception architecture transfer learning for garbage classification | |
| Lefaucheur et al. | Research and characterisation of blazar candidates among the Fermi/LAT 3FGL catalogue using multivariate classifications | |
| Das | Pattern Recognition using the Fuzzy c-means Technique | |
| CN107562938A (en) | A kind of law court intelligently tries method | |
| Kogler | Advances in Jet Substructure at the LHC | |
| CN103324758B (en) | A kind of news category method and system | |
| CN104951643B (en) | A kind of energy spectrum processing method suitable for neutron activation elementary analysis | |
| Abazov | Search for the standard model Higgs boson in $\ell\nu $+ jets final states in 9.7~ fb $^{-1} $ of $ p\bar {p} $ collisions with the D0 detector | |
| CN103310205B (en) | A kind of Handwritten Numeral Recognition Method and device | |
| Bai et al. | Learning single-particle mobility edges by a neural network based on data compression | |
| CN108107471A (en) | The acquisition methods and device of a kind of point of orientation first arrival data volume | |
| Athanasakos et al. | Is infrared-collinear safe information all you need for jet classification? | |
| CN107578445A (en) | Image discriminant region extracting method based on convolution characteristic spectrum | |
| CN104950000B (en) | A kind of close power spectrum finding method in peak position suitable for neutron activation elementary analysis | |
| CN106096647A (en) | A kind of RLID3 data classification method based on decision tree optimization rate | |
| Albert et al. | Model-independent search for neutrino sources with the ANTARES neutrino telescope | |
| CN104951452B (en) | A kind of similar power spectrum finding method in peak position suitable for neutron activation elementary analysis | |
| Kireyeu et al. | PHQMD Model for the Formation of Nuclear Clusters and Hypernuclei in Heavy Ion Collisions | |
| CN104951451B (en) | A kind of power spectrum construction of knowledge base method suitable for neutron activation elemental analysis | |
| CN104951453B (en) | A kind of consistent power spectrum finding method in peak position suitable for neutron activation elementary analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |