A data assimilation method, device, medium and product based on multiple linear regression model

Abstract

The present application discloses a data assimilation method, device, medium and product based on a multiple linear regression model, which relates to the field of data assimilation, using a DRI2001 carbon component analyzer to obtain first measurement concentration raw data and using a DRI2015 carbon component analyzer to obtain second measurement concentration raw data; according to the first measurement concentration raw data and the second measurement concentration raw data, a limited memory quasi-Newton optimization algorithm is used to solve the objective function group to determine the optimal assimilation coefficient group; according to the first measurement concentration raw data and the optimal assimilation coefficient group, a multiple linear regression model is used to determine the first measurement concentration assimilation data; the multiple linear regression model is the response relationship between the first measurement concentration raw data and the first measurement concentration assimilation data. The present application realizes the optimal assimilation of carbon component concentration data measured by the DRI2001 instrument and the DRI2015 instrument.

Description

Technical Field

[0001] The present application relates to the field of data assimilation, and in particular to a data assimilation method, device, medium and product based on a multiple linear regression model. Background Art

[0002] Carbonaceous components are PM in the ambient atmosphere. 2.5 An important component of PM 2.5 The concentration of PM2.5 is 20% to 50%, and its chemical composition is complex, mainly including organic carbon (OC) and elemental carbon (EC). Organic carbon is composed of hundreds of organic substances. There are significant differences in the pollution sources corresponding to different organic carbon components. There are both direct emissions and secondary conversions. Elemental carbon is carbon that exists in a single state, and its physical and chemical characteristics are related to the source emission type. In addition, the optical properties of different carbon-containing components are different, which can directly or indirectly affect global and regional climate change. At present, the most widely used and recognized mature carbon-containing component measurement method in various countries is thermal / optical analysis. Among them, the carbon component analyzer of the Desert Research Institute (DRI) in the United States is the most frequently used, widely used and most typical instrument. With the continuous deepening of scientific research and the iterative upgrading of science and technology, the DRI2001 carbon component analyzer commonly used in various countries more than 20 years ago has been discontinued. The DRI2015 carbon component analyzer is currently widely used to measure PM2.5. 2.5 However, there are obvious differences in the key components and detection technologies of DRI2001 and DRI2015 instruments. For example, the former uses He / Ne laser and FID as the light source and detector, while the latter uses 7-band modulated diode laser and NDIR as the light source and detector, respectively. This makes the carbon component measurement results of the two versions of instruments for the same group of samples significantly different, which has led to the differences in PM values ​​obtained by various countries in recent decades. 2.5 The data on medium carbon components cannot be effectively connected, which seriously hinders countries' scientific understanding of the evolution of carbonaceous aerosol pollution in recent decades. Summary of the invention

[0003] The purpose of this application is to provide a data assimilation method, device, medium and product based on a multiple linear regression model to achieve optimal assimilation of measurement data from two instruments.

[0004] To achieve the above objectives, this application provides the following solutions:

[0005] In a first aspect, the present application provides a data assimilation method based on a multiple linear regression model, comprising:

[0006] Obtain the first measured concentration raw data and the second measured concentration raw data; the first measured concentration raw data is measured using the DRI2001 carbon component analyzer; the second measured concentration raw data is measured using the DRI2015 carbon component analyzer; the first measured concentration raw data includes the PM to be detected 2.5 The first organic carbon first concentration, the second organic carbon first concentration, the third organic carbon first concentration, the fourth organic carbon first concentration, the first element carbon first concentration, the second element carbon first concentration and the third element carbon first concentration; the second measured concentration raw data includes the PM to be detected 2.5 a first organic carbon second concentration, a second organic carbon second concentration, a third organic carbon second concentration, a fourth organic carbon second concentration, a first element carbon second concentration, a second element carbon second concentration, and a third element carbon second concentration;

[0007] According to the first measured concentration raw data and the second measured concentration raw data, the objective function group is solved by using the limited memory quasi-Newton optimization algorithm to determine the optimal assimilation coefficient group; the objective function group includes the first objective function, the second objective function, the third objective function, the fourth objective function, the fifth objective function, the sixth objective function and the seventh objective function; the first objective function is an objective function that aims to minimize the relative standard deviation value between the assimilation data of the first organic carbon first concentration and the first organic carbon second concentration; the second objective function is an objective function that aims to minimize the relative standard deviation value between the assimilation data of the second organic carbon first concentration and the second organic carbon second concentration; the third objective function is an objective function that aims to minimize the relative standard deviation value between the assimilation data of the second organic carbon first concentration and the second organic carbon second concentration; The fourth objective function is an objective function that aims to minimize the relative standard deviation between the assimilation data of the first concentration of the fourth organic carbon and the second concentration of the fourth organic carbon; the fifth objective function is an objective function that aims to minimize the relative standard deviation between the assimilation data of the first concentration of the first element carbon and the second concentration of the first element carbon; the sixth objective function is an objective function that aims to minimize the relative standard deviation between the assimilation data of the first concentration of the second element carbon and the second concentration of the second element carbon; the seventh objective function is an objective function that aims to minimize the relative standard deviation between the assimilation data of the first concentration of the third element carbon and the second concentration of the third element carbon;

[0008] According to the first measured concentration original data and the optimal assimilation coefficient group, the first measured concentration assimilation data is determined using a multiple linear regression model; the multiple linear regression model is a response relationship between the first measured concentration original data and the first measured concentration assimilation data.

[0009] Optionally, according to the first measured concentration raw data and the second measured concentration raw data, solving the objective function group using a limited memory quasi-Newton optimization algorithm to determine the optimal assimilation coefficient group specifically includes:

[0010] According to the first organic carbon first concentration, the second organic carbon first concentration and the first organic carbon second concentration, with the first optimal assimilation coefficient being less than or equal to one as a constraint, a first objective function is fitted using a limited memory quasi-Newton optimization algorithm to determine a first optimal assimilation coefficient and a second optimal assimilation coefficient;

[0011] According to the first optimal assimilation coefficient, the first organic carbon first concentration, the second organic carbon first concentration, the third organic carbon first concentration and the second organic carbon second concentration, with the constraint that the sum of the second optimal assimilation coefficient and the third optimal assimilation coefficient is less than or equal to one, a limited memory quasi-Newton optimization algorithm is used to fit the second objective function to determine the third optimal assimilation coefficient and the fourth optimal assimilation coefficient;

[0012] According to the second optimal assimilation coefficient, the third optimal assimilation coefficient, the second organic carbon first concentration, the third organic carbon first concentration, the fourth organic carbon first concentration and the third organic carbon second concentration, with the constraint that the sum of the fourth optimal assimilation coefficient and the fifth optimal assimilation coefficient is less than or equal to one, the third objective function is fitted using a limited memory quasi-Newton optimization algorithm to determine the fifth optimal assimilation coefficient and the sixth optimal assimilation coefficient;

[0013] According to the fourth optimal assimilation coefficient, the fifth optimal assimilation coefficient, the third first organic carbon concentration, the fourth first organic carbon concentration, the first element carbon first concentration and the fourth second organic carbon concentration, with the constraint that the sum of the sixth optimal assimilation coefficient and the seventh optimal assimilation coefficient is less than or equal to one, the fourth objective function is fitted using a limited memory quasi-Newton optimization algorithm to determine the seventh optimal assimilation coefficient and the eighth optimal assimilation coefficient;

[0014] According to the sixth optimal assimilation coefficient, the seventh optimal assimilation coefficient, the fourth first organic carbon concentration, the first element carbon first concentration, the second element carbon first concentration and the first element carbon second concentration, with the sum of the eighth optimal assimilation coefficient and the ninth optimal assimilation coefficient being less than or equal to one as a constraint, a limited memory quasi-Newton optimization algorithm is used to fit the fifth objective function to determine the ninth optimal assimilation coefficient and the tenth optimal assimilation coefficient;

[0015] According to the eighth optimal assimilation coefficient, the ninth optimal assimilation coefficient, the tenth optimal assimilation coefficient, the first concentration of the first element carbon, the first concentration of the second element carbon, the first concentration of the third element carbon, the second concentration of the second element carbon and the second concentration of the third element carbon, with the sum of the tenth optimal assimilation coefficient and the eleventh optimal assimilation coefficient being less than or equal to one as a constraint, the limited memory quasi-Newton optimization algorithm is used to fit the sixth objective function and the seventh objective function to determine the eleventh optimal assimilation coefficient and the twelfth optimal assimilation coefficient;

[0016] The first optimal assimilation coefficient, the second optimal assimilation coefficient, the third optimal assimilation coefficient, the fourth optimal assimilation coefficient, the fifth optimal assimilation coefficient, the sixth optimal assimilation coefficient, the seventh optimal assimilation coefficient, the eighth optimal assimilation coefficient, the ninth optimal assimilation coefficient, the tenth optimal assimilation coefficient, the eleventh optimal assimilation coefficient and the twelfth optimal assimilation coefficient are determined as the optimal assimilation coefficient group.

[0017] Optionally, the objective function group is:

[0018] minδ1 2 =0.5*(a*OC1+b*OC2-OC1”) 2 ;

[0019] minδ2 2 =0.5*[(1-a)*OC1+c*OC2+d*OC3-OC2”] 2 ;

[0020] minδ3 2 =0.5*[(1-bc)*OC2+e*OC3+f*OC4-OC3”] 2 ;

[0021] minδ4 2 =0.5*[(1-de)*OC3+g*OC4+h*EC1-OC4”] 2 ;

[0022] minδ5 2 =0.5*[(1-fg)*OC4+i*EC1+j*EC2-EC1”] 2 ;

[0023] minδ6 2 =0.5*[(1-hi)*EC1+k*EC2+l*EC3-EC2”] 2 ;

[0024] minδ7 2=0.5*[(1-jk)*EC2+(1-l)*EC3-EC3”] 2 ;

[0025] Among them, δ1 is the relative standard deviation value between the assimilation data of the first organic carbon concentration and the second organic carbon concentration; δ2 is the relative standard deviation value between the assimilation data of the second organic carbon concentration and the second organic carbon concentration; δ3 is the relative standard deviation value between the assimilation data of the third organic carbon concentration and the second organic carbon concentration; δ4 is the relative standard deviation value between the assimilation data of the fourth organic carbon concentration and the second organic carbon concentration; δ5 is the relative standard deviation value between the assimilation data of the first element carbon concentration and the second element carbon concentration; δ6 is the relative standard deviation value between the assimilation data of the second element carbon concentration and the second element carbon concentration; δ7 is the relative standard deviation value between the assimilation data of the third element carbon concentration and the second element carbon concentration; a is the first optimal assimilation coefficient; b is the second optimal assimilation coefficient; c is the third optimal assimilation coefficient; d is the fourth optimal assimilation coefficient. coefficient; e is the fifth optimal assimilation coefficient; f is the sixth optimal assimilation coefficient; g is the seventh optimal assimilation coefficient; h is the eighth optimal assimilation coefficient; i is the ninth optimal assimilation coefficient; j is the tenth optimal assimilation coefficient; k is the eleventh optimal assimilation coefficient; l is the twelfth optimal assimilation coefficient; OC1 is the first concentration of the first organic carbon; OC2 is the first concentration of the second organic carbon; OC3 is the first concentration of the third organic carbon; OC4 is the first concentration of the fourth organic carbon; EC1 is the first concentration of the first element carbon; EC2 is the first concentration of the second element carbon; EC3 is the first concentration of the third element carbon; OC1” is the second concentration of the first organic carbon; OC2” is the second concentration of the second organic carbon; OC3” is the second concentration of the third organic carbon; OC4” is the second concentration of the fourth organic carbon; EC1” is the second concentration of the first element carbon; EC2” is the second concentration of the second element carbon; EC3” is the second concentration of the third element carbon.

[0026] Optionally, the multiple linear regression model is:

[0027] OC1' = a*OC1 + b*OC2;

[0028] OC2'=(1-a)*OC1+c*OC2+d*OC3;

[0029] OC3'=(1-bc)*OC2+e*OC3+f*OC4;

[0030] OC4'=(1-de)*OC3+g*OC4+h*EC1;

[0031] EC1'=(1-fg)*OC4+i*EC1+j*EC2;

[0032] EC2'=(1-hi)*EC1+k*EC2+l*EC3;

[0033] EC3'=(1-jk)*EC2+(1-l)*EC3;

[0034] Among them, OC1' is the first organic carbon first concentration assimilation data; OC2' is the second organic carbon first concentration assimilation data; OC3' is the third organic carbon first concentration assimilation data; OC4' is the fourth organic carbon first concentration assimilation data; EC1' is the first element carbon first concentration assimilation data; EC2' is the second element carbon first concentration assimilation data; EC3' is the third element carbon first concentration assimilation data; OC1 is the first organic carbon first concentration; OC2 is the second organic carbon first concentration; OC3 is the third organic carbon first concentration; OC4 is the fourth The first concentration of organic carbon; EC1 is the first concentration of the first element carbon; EC2 is the first concentration of the second element carbon; EC3 is the first concentration of the third element carbon; a is the first optimal assimilation coefficient; b is the second optimal assimilation coefficient; c is the third optimal assimilation coefficient; d is the fourth optimal assimilation coefficient; e is the fifth optimal assimilation coefficient; f is the sixth optimal assimilation coefficient; g is the seventh optimal assimilation coefficient; h is the eighth optimal assimilation coefficient; i is the ninth optimal assimilation coefficient; j is the tenth optimal assimilation coefficient; k is the eleventh optimal assimilation coefficient; l is the twelfth optimal assimilation coefficient.

[0035] Optionally, the DRI2001 carbon component analyzer uses a 633nm He / Ne laser and a FID detector to measure the first measurement concentration raw data; the DRI2015 carbon component analyzer uses a 635nm diode laser and a NDIR detector to measure the second measurement concentration raw data.

[0036] In a second aspect, the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement any of the above-described data assimilation methods based on a multiple linear regression model.

[0037] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-mentioned data assimilation methods based on a multiple linear regression model.

[0038] In a fourth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned data assimilation methods based on a multiple linear regression model.

[0039] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0040] The present application provides a data assimilation method, device, medium and product based on a multiple linear regression model, which obtains first measured concentration raw data and second measured concentration raw data; the first measured concentration raw data is measured using a DRI2001 carbon component analyzer; the second measured concentration raw data is measured using a DRI2015 carbon component analyzer; according to the first measured concentration raw data and the second measured concentration raw data, the objective function group is solved using a limited memory quasi-Newton optimization algorithm to determine an optimal assimilation coefficient group; according to the first measured concentration raw data and the optimal assimilation coefficient group, the first measured concentration assimilation data is determined using a multiple linear regression model; the multiple linear regression model is the response relationship between the first measured concentration raw data and the first measured concentration assimilation data. The present application realizes the optimal assimilation of carbon component concentration data measured by the DRI2001 instrument and the DRI2015 instrument. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 A flowchart of a data assimilation method based on a multiple linear regression model provided in one embodiment of the present application;

[0043] Figure 2 A flowchart of the data assimilation method based on the multiple linear regression model in practical application of the present application;

[0044] Figure 3 PM measurements for the DRI2001 instrument 2.5 Relative standard deviation curves of area concentrations of TC, OC, EC, OC1, OC2, OC3, OC4, EC1, EC2 and EC3 in parallel samples;

[0045] Figure 4 Measuring PM for the DRI2015 instrument 2.5 Relative standard deviation curves of area concentrations of TC, OC, EC, OC1, OC2, OC3, OC4, EC1, EC2 and EC3 in parallel samples;

[0046] Figure 5 It is a comparison curve between the original concentration value and the concentration after assimilation;

[0047] Figure 6 It is the relationship curve between relative standard deviation and frequency of occurrence;

[0048] Figure 7 This is an application environment diagram of a data assimilation method based on a multiple linear regression model in one embodiment of the present application;

[0049] Figure 8 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0051] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0052] In an exemplary embodiment, Figure 1 and Figure 2 As shown, a data assimilation method based on a multiple linear regression model is provided, including:

[0053] S1: Obtain the first measured concentration raw data and the second measured concentration raw data; the first measured concentration raw data is measured using a DRI2001 carbon component analyzer; the second measured concentration raw data is measured using a DRI2015 carbon component analyzer; the first measured concentration raw data includes the PM to be detected 2.5 The first organic carbon first concentration, the second organic carbon first concentration, the third organic carbon first concentration, the fourth organic carbon first concentration, the first element carbon first concentration, the second element carbon first concentration and the third element carbon first concentration; the second measured concentration raw data includes the PM to be detected 2.5 The first organic carbon second concentration, the second organic carbon second concentration, the third organic carbon second concentration, the fourth organic carbon second concentration, the first element carbon second concentration, the second element carbon second concentration and the third element carbon second concentration.

[0054] The DRI2001 carbon component analyzer uses a 633nm He / Ne laser and a FID detector to measure the first measurement concentration raw data; the DRI2015 carbon component analyzer uses a 635nm diode laser and a NDIR detector to measure the second measurement concentration raw data.

[0055] In practical applications, the carbonaceous component concentration is first measured: PM 2.5 The specific process of carbon component concentration is as follows: the DRI instrument adopts the IMPROVE heating program for measurement, so the concentrations of different carbonaceous components are defined by the measurement process of different heating sections. For example, heating at 140℃, 280℃, 480℃, 580℃, 740℃ and 840℃ can measure the concentrations of OC1, OC2, OC3, OC4, EC1, EC2 and EC3 respectively.

[0056] In view of the obvious differences in key components and detection technologies between DRI2001 and DRI2015 instruments, DRI2001 uses a 633nm He / Ne laser and FID detector to measure the concentrations of OC1, OC2, OC3, OC4, EC1, EC2 and EC3, while DRI2015 uses a 635nm diode laser and NDIR detector to measure the concentrations of the above carbon components. Therefore, there are significant differences in the concentrations of OC1, OC2, OC3, OC4, EC1, EC2 and EC3 measured by the two versions of the instruments.

[0057] In practical applications, before obtaining the optimized data for DRI2001 measurement concentration, it is necessary to first construct the response relationship between the original data of different carbonaceous component concentrations measured by the DRI2001 instrument and the optimized data, that is, the multiple linear regression model and the objective function group.

[0058] The multiple linear regression model is:

[0059] OC1'=a*OC1+b*OC2.

[0060] OC2'=(1-a)*OC1+c*OC2+d*OC3.

[0061] OC3'=(1-bc)*OC2+e*OC3+f*OC4.

[0062] OC4'=(1-de)*OC3+g*OC4+h*EC1.

[0063] EC1'=(1-fg)*OC4+i*EC1+j*EC2.

[0064] EC2'=(1-hi)*EC1+k*EC2+l*EC3.

[0065] EC3'=(1-jk)*EC2+(1-l)*EC3.

[0066] Among them, OC1' is the first organic carbon first concentration assimilation data; OC2' is the second organic carbon first concentration assimilation data; OC3' is the third organic carbon first concentration assimilation data; OC4' is the fourth organic carbon first concentration assimilation data; EC1' is the first element carbon first concentration assimilation data; EC2' is the second element carbon first concentration assimilation data; EC3' is the third element carbon first concentration assimilation data.

[0067] a, b, c, d, e, f, g, h, i, j, k and l are optimization coefficients, all ≥ 0.01, and a≤1.00, b+c≤1.00, d+e≤1.00, f+g≤1.00, h+i≤1.00, j+k≤1.00, l≤1.00.

[0068] The optimized data OC1', OC2', OC3', OC4', EC1', EC2' and EC3' of OC1, OC2, OC3, OC4, EC1, EC2 and EC3 measured by DRI2001 were respectively matched with the original data of OC1", OC2", OC3", OC4", EC1", EC2" and EC3" measured by DRI2015, and the calculation formula of the relative standard deviation between the two components was established as follows:

[0069] δ1 2 =0.5*(a*OC1+b*OC2-OC1”) 2 .

[0070] δ2 2 =0.5*[(1-a)*OC1+c*OC2+d*OC3-OC2”] 2 .

[0071] δ3 2 =0.5*[(1-bc)*OC2+e*OC3+f*OC4-OC3”] 2 .

[0072] δ4 2 =0.5*[(1-de)*OC3+g*OC4+h*EC1-OC4”] 2 .

[0073] δ5 2 =0.5*[(1-fg)*OC4+i*EC1+j*EC2-EC1”] 2 .

[0074] δ6 2 =0.5*[(1-hi)*EC1+k*EC2+l*EC3-EC2”] 2 .

[0075] δ72 =0.5*[(1-jk)*EC2+(1-l)*EC3-EC3”] 2 .

[0076] δ1, δ2, δ3, δ4, δ5, δ6 and δ7 represent the relative standard deviations between OC1” and OC1’, OC2” and OC2’, OC3” and OC3’, OC4” and OC4’, EC1” and EC1’, EC2” and EC2’, EC3” and EC3’, respectively.

[0077] The objective function group is:

[0078] minδ1 2 =0.5*(a*OC1+b*OC2-OC1”) 2 .

[0079] minδ2 2 =0.5*[(1-a)*OC1+c*OC2+d*OC3-OC2”] 2 .

[0080] minδ3 2 =0.5*[(1-bc)*OC2+e*OC3+f*OC4-OC3”] 2 .

[0081] minδ4 2 =0.5*[(1-de)*OC3+g*OC4+h*EC1-OC4”] 2 .

[0082] minδ5 2 =0.5*[(1-fg)*OC4+i*EC1+j*EC2-EC1”] 2 .

[0083] minδ6 2 =0.5*[(1-hi)*EC1+k*EC2+l*EC3-EC2”] 2 .

[0084] minδ7 2 =0.5*[(1-jk)*EC2+(1-l)*EC3-EC3”] 2 .

[0085] Among them, δ1 is the relative standard deviation value between the assimilation data of the first organic carbon concentration and the second organic carbon concentration; δ2 is the relative standard deviation value between the assimilation data of the second organic carbon concentration and the second organic carbon concentration; δ3 is the relative standard deviation value between the assimilation data of the third organic carbon concentration and the second organic carbon concentration; δ4 is the relative standard deviation value between the assimilation data of the fourth organic carbon concentration and the second organic carbon concentration; δ5 is the relative standard deviation value between the assimilation data of the first element carbon concentration and the second element carbon concentration; δ6 is the relative standard deviation value between the assimilation data of the second element carbon concentration and the second element carbon concentration; δ7 is the relative standard deviation value between the assimilation data of the third element carbon concentration and the second element carbon concentration; a is the first optimal assimilation coefficient; b is the second optimal assimilation coefficient; c is the third optimal assimilation coefficient; d is the fourth optimal assimilation coefficient. coefficient; e is the fifth optimal assimilation coefficient; f is the sixth optimal assimilation coefficient; g is the seventh optimal assimilation coefficient; h is the eighth optimal assimilation coefficient; i is the ninth optimal assimilation coefficient; j is the tenth optimal assimilation coefficient; k is the eleventh optimal assimilation coefficient; l is the twelfth optimal assimilation coefficient; OC1 is the first concentration of the first organic carbon; OC2 is the first concentration of the second organic carbon; OC3 is the first concentration of the third organic carbon; OC4 is the first concentration of the fourth organic carbon; EC1 is the first concentration of the first element carbon; EC2 is the first concentration of the second element carbon; EC3 is the first concentration of the third element carbon; OC1” is the second concentration of the first organic carbon; OC2” is the second concentration of the second organic carbon; OC3” is the second concentration of the third organic carbon; OC4” is the second concentration of the fourth organic carbon; EC1” is the second concentration of the first element carbon; EC2” is the second concentration of the second element carbon; EC3” is the second concentration of the third element carbon.

[0086] S2: According to the first measured concentration raw data and the second measured concentration raw data, the objective function group is solved by using the limited memory quasi-Newton optimization algorithm to determine the optimal assimilation coefficient group; the objective function group includes the first objective function, the second objective function, the third objective function, the fourth objective function, the fifth objective function, the sixth objective function and the seventh objective function; the first objective function is an objective function that aims to minimize the relative standard deviation between the assimilation data of the first organic carbon first concentration and the first organic carbon second concentration; the second objective function is an objective function that aims to minimize the relative standard deviation between the assimilation data of the second organic carbon first concentration and the second organic carbon second concentration; the third objective function is an objective function that aims to minimize the relative standard deviation between the assimilation data of the second organic carbon first concentration and the second organic carbon second concentration; The objective function aims to minimize the relative standard deviation between the assimilated data of the first concentration and the second concentration of the third organic carbon; the fourth objective function aims to minimize the relative standard deviation between the assimilated data of the first concentration of the fourth organic carbon and the second concentration of the fourth organic carbon; the fifth objective function aims to minimize the relative standard deviation between the assimilated data of the first concentration of the first element carbon and the second concentration of the first element carbon; the sixth objective function aims to minimize the relative standard deviation between the assimilated data of the first concentration of the second element carbon and the second concentration of the second element carbon; the seventh objective function aims to minimize the relative standard deviation between the assimilated data of the first concentration of the third element carbon and the second concentration of the third element carbon.

[0087] Secondly, based on the original concentration data of OC1, OC2, OC3, OC4, EC1, EC2, EC3 and OC1”, OC2”, OC3”, OC4”, EC1”, EC2” and EC3” obtained from the above measurements, the optimal assimilation coefficients of the original concentration data measured by DRI2001 and DRI2015 were obtained: by optimizing the minimum values ​​of the seven objective functions in turn, and then obtaining the optimal assimilation coefficients a, b, c, d, e, f, g, h, i, j, k and l.

[0088] The optimal assimilation coefficient of the raw data of DRI2001 and DRI2015 concentration measurements is obtained in the following way:

[0089] According to the first organic carbon first concentration, the second organic carbon first concentration and the first organic carbon second concentration, with the first optimal assimilation coefficient being less than or equal to one as a constraint, a limited memory quasi-Newton optimization algorithm is used to fit the first objective function to determine the first optimal assimilation coefficient and the second optimal assimilation coefficient.

[0090] Through the limited memory quasi-Newton optimization algorithm, the minimum value of the first objective function is first optimized to obtain the optimal assimilation coefficients a and b.

[0091] According to the first optimal assimilation coefficient, the first organic carbon first concentration, the second organic carbon first concentration, the third organic carbon first concentration and the second organic carbon second concentration, with the constraint that the sum of the second optimal assimilation coefficient and the third optimal assimilation coefficient is less than or equal to one, the second objective function is fit using a limited memory quasi-Newton optimization algorithm to determine the third optimal assimilation coefficient and the fourth optimal assimilation coefficient.

[0092] Substitute the a value obtained above into the second objective function, and under the constraint condition of b+c≤1.00, perform optimal fitting on the minimum value of the second objective function to obtain the optimal assimilation coefficients c and d.

[0093] According to the second optimal assimilation coefficient, the third optimal assimilation coefficient, the second organic carbon first concentration, the third organic carbon first concentration, the fourth organic carbon first concentration and the third organic carbon second concentration, with the constraint that the sum of the fourth optimal assimilation coefficient and the fifth optimal assimilation coefficient is less than or equal to one, the third objective function is fit using a limited memory quasi-Newton optimization algorithm to determine the fifth optimal assimilation coefficient and the sixth optimal assimilation coefficient.

[0094] Substitute the b and c values ​​obtained above into the third objective function, and optimize the minimum value of the third objective function under the constraint condition of d+e≤1.00 to obtain the optimal assimilation coefficients e and f.

[0095] According to the fourth optimal assimilation coefficient, the fifth optimal assimilation coefficient, the third first organic carbon concentration, the fourth first organic carbon concentration, the first element carbon first concentration and the fourth second organic carbon concentration, with the constraint that the sum of the sixth optimal assimilation coefficient and the seventh optimal assimilation coefficient is less than or equal to one, the fourth objective function is fit using a limited memory quasi-Newton optimization algorithm to determine the seventh optimal assimilation coefficient and the eighth optimal assimilation coefficient.

[0096] Substitute the above-obtained d and e values ​​into the fourth objective function, and under the constraint condition of f+g≤1.00, perform optimal fitting on the minimum value of the fourth objective function to obtain the optimal assimilation coefficients g and h values.

[0097] According to the sixth optimal assimilation coefficient, the seventh optimal assimilation coefficient, the fourth first organic carbon concentration, the first element carbon first concentration, the second element carbon first concentration and the first element carbon second concentration, with the constraint that the sum of the eighth optimal assimilation coefficient and the ninth optimal assimilation coefficient is less than or equal to one, the fifth objective function is fit using a limited memory quasi-Newton optimization algorithm to determine the ninth optimal assimilation coefficient and the tenth optimal assimilation coefficient.

[0098] Substitute the f and g values ​​obtained above into the fifth objective function, and optimize the minimum value of the fifth objective function under the constraint condition of h+i≤1.00 to obtain the optimal assimilation coefficients i and j.

[0099] According to the eighth optimal assimilation coefficient, the ninth optimal assimilation coefficient, the tenth optimal assimilation coefficient, the first element carbon first concentration, the second element carbon first concentration, the third element carbon first concentration, the second element carbon second concentration and the third element carbon second concentration, with the constraint that the sum of the tenth optimal assimilation coefficient and the eleventh optimal assimilation coefficient is less than or equal to one, the limited memory quasi-Newton optimization algorithm is used to fit the sixth objective function and the seventh objective function to determine the eleventh optimal assimilation coefficient and the twelfth optimal assimilation coefficient.

[0100] The h, i and j values ​​obtained above are substituted into the sixth objective function and the seventh objective function respectively, and the minimum value of the sum of the sixth objective function and the seventh objective function is optimally fitted under the constraint condition of j+k≤1.00 to obtain the optimal assimilation coefficients k and l.

[0101] The first optimal assimilation coefficient, the second optimal assimilation coefficient, the third optimal assimilation coefficient, the fourth optimal assimilation coefficient, the fifth optimal assimilation coefficient, the sixth optimal assimilation coefficient, the seventh optimal assimilation coefficient, the eighth optimal assimilation coefficient, the ninth optimal assimilation coefficient, the tenth optimal assimilation coefficient, the eleventh optimal assimilation coefficient and the twelfth optimal assimilation coefficient are determined as the optimal assimilation coefficient group.

[0102] When the optimization fitting is specifically solved, a, b, c, d, e, f, g, h, i, j, k and l are between 0.01 and 1.00. Specific implementation methods: Python, Matlab and other computer programming languages ​​can be used for data processing and iterative solving.

[0103] The limited memory quasi-Newton optimization algorithm is an optimization algorithm that uses the first 3 to 20 iterations to approximate the minimum value of the objective function. This method combines the quasi-Newton optimization algorithm and the conjugate direction optimization algorithm to perform limited memory loop optimization, which can greatly shorten the time to obtain the optimal solution and has the characteristics of low iteration cost, fast solution speed and simple programming.

[0104] Regarding the calculation formula of the limited memory quasi-Newton optimization algorithm, this application uses the example of optimizing the minimum value of the first objective function to obtain the optimal assimilation coefficients a and b as an example to explain as follows. Considering that after substituting the a value obtained by optimizing the first objective function into the second objective function, only c and d are unknowns in the second objective function, which is consistent with the solution algorithm of the first objective function; by analogy, the algorithms for solving the optimization coefficients of the third objective function, the fourth objective function, the fifth objective function, the sixth objective function, and the seventh objective function are also consistent with the first objective function.

[0105] First, this application uses the minf(x) form of optimization solution mathematical model instead of minδ 2 To explain, X1 represents the above a, X2 represents the above b, f(x) = 0.5 (αX1 + βX2 + γ) 2 , where α, β and γ are known numbers, and x=(x1,x2,…x n )T is the variable of this optimization solution, and its value is between 0.01 and 1.00; the objective function f is R n →R is a continuously differentiable convex function, and X* is the optimal solution. The basic idea of ​​the above iteration is: first, two initial points x0 are selected, where the first initial point x0 is used to fix the value of X1, and the second initial point x0 is used to solve minf(x), that is, on the basis of fixing the value of X1, the second initial point starts from x0 and steps with a length of λ k And the f(x) value is iterated with a downward trend, and a series of points x1, x2, ···, x are obtained in turn. k ,···, the point series is {x k}, That is the 0th optimal solution of minf(x); secondly, the first initial point x0 is calculated with a step size of λ k The value of X1 is also fixed k times in turn, and the second initial point x0 is still fixed using the above iterative rule to obtain the kth optimal solution of minf(x); finally, the minimum value among 0 to k times minf(x) is selected, and its corresponding X1* and X2* values ​​are the optimal solutions of the objective function minf(x).

[0106] The algorithm for solving the 0th to kth optimal solution of minf(x) in this application is based on the quasi-Newton optimization algorithm with limited memory optimization. Among them, the most important step of the quasi-Newton optimization algorithm is to solve the "simulation" matrix H of the Hessian matrix at each iteration k+1 , and then calculate the iteration direction and bring it into the iteration formula to find the point sequence {x k Here, we assume that R n →R is continuous and at least twice differentiable, then f(x) is k+1 is twice differentiable at x.k+1 Do a second-order Taylor expansion and take the derivative:

[0107]

[0108] Let: x = x k , σ k =x k+1 -x k , have to:

[0109] y k =H k+1 σ k .

[0110] Among them, σ k is the step length; y k is f(x) at x k+1 and x k The gradient difference at is f(x) at x k+1 The gradient at is f(x) at x k The gradient at .

[0111] Given the matrix H k+1 It must be a symmetric positive definite matrix to ensure that the value of f(x) is on a downward trend, so it must satisfy In addition, I represents the identity matrix. Then the only solution of the above formula is:

[0112]

[0113] Compared with the quasi-Newton optimization algorithm, each correction of the limited memory quasi-Newton optimization algorithm is based on the initially given symmetric positive definite matrix No need to remember H k , so only the initial matrix needs to be stored and m+1 vector groups You can correct H k+1 , m is a non-negative integer, which makes the algorithm occupy less resources and has high operating efficiency. The specific operating steps are as follows:

[0114] Step 1: Take the initial point x0∈R n , H0∈R n×n , accuracy θ>0, let k=0.

[0115] Step 2: If The algorithm terminates and the solution obtained is x k , otherwise, let

[0116] Step 3: Determine the step size λ according to the exact line search criterionk , the criterion takes the step size And satisfy the orthogonality Among them, λ is the step size.

[0117] Step 4: Let x k+1 =x k +λd k ,calculate when Time k+1 is the optimal solution, otherwise proceed to the next step.

[0118] Step 5: Order H is calculated based on the following formula k+1 :

[0119]

[0120]

[0121] Step 6: Let k=k+1 and go to step 2.

[0122] S3: Determine the first measured concentration assimilated data using a multiple linear regression model based on the first measured concentration original data and the optimal assimilation coefficient group; the multiple linear regression model is a response relationship between the first measured concentration original data and the first measured concentration assimilated data.

[0123] The optimal assimilation coefficients a, b, c, d, e, f, g, h, i, j, k and l values ​​obtained above were respectively substituted into the above-mentioned multiple linear regression model to obtain the assimilated OC1', OC2', OC3', OC4', EC1', EC2' and EC3' values. Specific embodiment:

[0125] The case study was based on about 200 ambient air PM values ​​in urban and suburban areas of Chongqing from 2013 to 2021. 2.5 The samples were used as the implementation objects, and the PM was measured by DRI2001 and DRI2015 instruments respectively. 2.5 The carbon component concentration in the medium is obtained by using the multiple linear regression model method to perform minimum error assimilation on the data measured by the two instruments to obtain the optimized concentration of the carbon component measured by the DRI2001 instrument.

[0126] (1) Assessment of uncertainty in instrument measurement data.

[0127] First, this implementation case evaluates the uncertainty of carbon component data of parallel samples measured by DRI2001 and DRI2015 instruments to ensure that the measurement data of the two instruments are reasonable and reliable. Figure 3 and Figure 4As shown, the horizontal axis is the carbon component area concentration (μg·cm -2 ), the ordinate is the absolute value of the relative standard deviation |RV|:

[0128]

[0129] The blue dotted line represents the minimum detection limit of 10 times the TC concentration; the red solid line represents the uncertainty of the instrument at the 95% confidence level. Its |RV| corresponds to the relative error value CV of the parallel sample measurement is less than the minimum detection limit MDL of the instrument. The calculation formula is as follows:

[0130]

[0131] See also Figure 3 and Figure 4 In this implementation case, DRI2001 and DRI2015 instruments measure PM 2.5 The relative standard deviations of the area concentrations of TC, OC, EC, OC1, OC2, OC3, OC4, EC1, EC2 and EC3 in the parallel samples are all within the uncertainty of the instrument at the 95% confidence level, indicating that the data of all carbon components measured by the two instruments are reasonable and reliable. In addition, among all carbonaceous components, the relative standard deviations of the area concentrations of OC1 and EC3 are relatively large, while the relative standard deviations of the area concentrations of TC, OC2 and OC3 are relatively small.

[0132] (2) Comparison and analysis of data optimization results.

[0133] See also Figure 5 In this implementation case, the slopes between the initial concentrations of TC, OC, EC, OC1, OC2, OC3, OC4, EC1, EC2, and EC3 measured by the DRI2001 and DRI2015 instruments were 1.03, 1.04, 1.00, 2.42, 0.97, 0.98, 0.89, 1.09, and 0.39, respectively, and the corresponding correlation coefficients R 2 The carbon components measured by the DRI2001 instrument were optimized by the multiple linear regression model method, and the obtained TC, OC, EC, OC1, OC2, OC3, OC4, EC1, EC2 and EC3 optimized concentrations and the initial concentrations of the corresponding carbon components measured by the DRI2015 instrument were significantly better than the results before optimization, that is, the slope between the optimized concentration and the initial concentration of the carbon components was closer to 1.00, which were 1.01, 1.00, 1.04, 0.95, 1.00, 1.00, 1.04 and 0.86, respectively, and the corresponding correlation coefficients R 2The values ​​are larger, namely 0.97, 0.96, 0.91, 0.66, 0.95, 0.93, 0.87, 0.94 and 0.82 respectively.

[0134] See also Figure 6 and Table 1, in this embodiment, the relative standard deviations between the initial concentrations of TC, OC, EC, OC1, OC2, OC3, OC4, EC1, EC2 and EC3 measured by DRI2001 and DRI2015 instruments were 4.7±3.7%, 5.7±4.6%, 10.2±8.4%, 62.3±17.0%, 7.9±6.4%, 7.3±5.5%, 10.8±9.4%, 11.6±9.7%, 50.2±26.2% and 121.5±37.2%, respectively. The carbon components measured by the DRI2001 instrument were optimized by the multiple linear regression model method, and the relative standard deviations between the optimized concentrations of TC, OC, EC, OC1, OC2, OC3, OC4, EC1, EC2 and EC3 and the initial concentrations of the corresponding carbon components measured by the DRI2015 instrument were 4.3±3.4%, 4.4±3.8%, 10.5±7.8%, 15.4±13.2%, 5.2±4.2%, 7.1±4.7%, 7.6±7.2%, 10.1±7.6%, 25.8±19.5% and 84.5±55.3%, respectively, which were significantly smaller than the results before optimization (except EC), and the frequency of relative standard deviations of each carbon component after optimization was more concentrated than the results before optimization.

[0135] Table 1 Relative standard deviation comparison table

[0136]

[0137]

[0138] In summary, in this implementation case, the multiple linear regression model method has a very positive effect on assimilating the carbon component data measured by the DRI2001 and DRI2015 instruments.

[0139] The multiple linear regression model method based on the measurement data of different carbon components provided in this application can realize the PM measurement over decades using DRI instruments. 2.5 The effective connection of the carbonaceous component concentration data will help to accurately understand the pollution evolution of carbonaceous aerosols in various countries in recent decades.

[0140] The present application also provides a data processing system device, including a computer device of a memory and a processor, wherein the memory and the processor communicate with each other through a bus, wherein the memory stores program instructions of a data assimilation method based on a multiple linear regression model, and the processor executes the above-mentioned computer program stored in the memory, and its data processing terminal can output the processed information data.

[0141] The data assimilation method based on the multiple linear regression model provided in the embodiment of the present application can be applied to Figure 7 In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the first measurement concentration raw data and the second measurement concentration raw data to be processed to the server 104. After the server 104 receives the first measurement concentration raw data and the second measurement concentration raw data to be processed, for the first measurement concentration raw data and the second measurement concentration raw data to be processed, the server 104 uses the DRI2001 carbon component analyzer to obtain the first measurement concentration raw data and the DRI2015 carbon component analyzer to obtain the second measurement concentration raw data; according to the first measurement concentration raw data and the second measurement concentration raw data, the limited memory quasi-Newton optimization algorithm is used to solve the objective function group and determine the optimal assimilation coefficient group; according to the first measurement concentration raw data and the optimal assimilation coefficient group, the multiple linear regression model is used to determine the first measurement concentration assimilation data. The server 104 can feed back the obtained first measurement concentration assimilation data to the terminal 102. In addition, in some embodiments, the data assimilation method based on the multiple linear regression model can also be implemented independently by the server 104 or the terminal 102. For example, the terminal 102 can directly perform data assimilation processing on the first measurement concentration raw data and the second measurement concentration raw data to be processed, or the server 104 can obtain the first measurement concentration raw data and the second measurement concentration raw data to be processed from the data storage system, and perform data assimilation processing on the first measurement concentration raw data and the second measurement concentration raw data to be processed.

[0142] The terminal 102 may be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart vehicle-mounted devices, etc. The portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or may be a cloud server.

[0143] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 8 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store carbonaceous component concentration data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a data assimilation method based on a multiple linear regression model is implemented.

[0144] Those skilled in the art will understand that Figure 8 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0145] In an exemplary embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0146] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0147] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0148] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0149] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0150] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0151] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0152] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A data assimilation method based on a multiple linear regression model, characterized in that: include: Obtain the first measured concentration raw data and the second measured concentration raw data; the first measured concentration raw data is measured using the DRI2001 carbon component analyzer; the second measured concentration raw data is measured using the DRI2015 carbon component analyzer; the first measured concentration raw data includes the PM to be detected 2.5 The first organic carbon first concentration, the second organic carbon first concentration, the third organic carbon first concentration, the fourth organic carbon first concentration, the first element carbon first concentration, the second element carbon first concentration and the third element carbon first concentration; the second measured concentration raw data includes the PM to be detected 2.5 a first organic carbon second concentration, a second organic carbon second concentration, a third organic carbon second concentration, a fourth organic carbon second concentration, a first element carbon second concentration, a second element carbon second concentration, and a third element carbon second concentration; According to the first measured concentration raw data and the second measured concentration raw data, the objective function group is solved by using the limited memory quasi-Newton optimization algorithm to determine the optimal assimilation coefficient group; the objective function group includes the first objective function, the second objective function, the third objective function, the fourth objective function, the fifth objective function, the sixth objective function and the seventh objective function; the first objective function is an objective function that aims to minimize the relative standard deviation value between the assimilation data of the first organic carbon first concentration and the first organic carbon second concentration; the second objective function is an objective function that aims to minimize the relative standard deviation value between the assimilation data of the second organic carbon first concentration and the second organic carbon second concentration; the third objective function is an objective function that aims to minimize the relative standard deviation value between the assimilation data of the second organic carbon first concentration and the second organic carbon second concentration; The fourth objective function is an objective function that aims to minimize the relative standard deviation between the assimilation data of the first concentration of the fourth organic carbon and the second concentration of the fourth organic carbon; the fifth objective function is an objective function that aims to minimize the relative standard deviation between the assimilation data of the first concentration of the first element carbon and the second concentration of the first element carbon; the sixth objective function is an objective function that aims to minimize the relative standard deviation between the assimilation data of the first concentration of the second element carbon and the second concentration of the second element carbon; the seventh objective function is an objective function that aims to minimize the relative standard deviation between the assimilation data of the first concentration of the third element carbon and the second concentration of the third element carbon; According to the first measured concentration original data and the optimal assimilation coefficient group, the first measured concentration assimilation data is determined using a multiple linear regression model; the multiple linear regression model is a response relationship between the first measured concentration original data and the first measured concentration assimilation data.

2. The data assimilation method based on multiple linear regression model according to claim 1, characterized in that: According to the first measured concentration raw data and the second measured concentration raw data, the objective function group is solved by using a limited memory quasi-Newton optimization algorithm to determine an optimal assimilation coefficient group, specifically including: According to the first organic carbon first concentration, the second organic carbon first concentration and the first organic carbon second concentration, with the first optimal assimilation coefficient being less than or equal to one as a constraint, a first objective function is fitted using a limited memory quasi-Newton optimization algorithm to determine a first optimal assimilation coefficient and a second optimal assimilation coefficient; According to the first optimal assimilation coefficient, the first organic carbon first concentration, the second organic carbon first concentration, the third organic carbon first concentration and the second organic carbon second concentration, with the constraint that the sum of the second optimal assimilation coefficient and the third optimal assimilation coefficient is less than or equal to one, a limited memory quasi-Newton optimization algorithm is used to fit the second objective function to determine the third optimal assimilation coefficient and the fourth optimal assimilation coefficient; According to the second optimal assimilation coefficient, the third optimal assimilation coefficient, the second organic carbon first concentration, the third organic carbon first concentration, the fourth organic carbon first concentration and the third organic carbon second concentration, with the constraint that the sum of the fourth optimal assimilation coefficient and the fifth optimal assimilation coefficient is less than or equal to one, the third objective function is fitted using a limited memory quasi-Newton optimization algorithm to determine the fifth optimal assimilation coefficient and the sixth optimal assimilation coefficient; According to the fourth optimal assimilation coefficient, the fifth optimal assimilation coefficient, the third first organic carbon concentration, the fourth first organic carbon concentration, the first element carbon first concentration and the fourth second organic carbon concentration, with the constraint that the sum of the sixth optimal assimilation coefficient and the seventh optimal assimilation coefficient is less than or equal to one, the fourth objective function is fitted using a limited memory quasi-Newton optimization algorithm to determine the seventh optimal assimilation coefficient and the eighth optimal assimilation coefficient; According to the sixth optimal assimilation coefficient, the seventh optimal assimilation coefficient, the fourth first organic carbon concentration, the first element carbon first concentration, the second element carbon first concentration and the first element carbon second concentration, with the sum of the eighth optimal assimilation coefficient and the ninth optimal assimilation coefficient being less than or equal to one as a constraint, a limited memory quasi-Newton optimization algorithm is used to fit the fifth objective function to determine the ninth optimal assimilation coefficient and the tenth optimal assimilation coefficient; According to the eighth optimal assimilation coefficient, the ninth optimal assimilation coefficient, the tenth optimal assimilation coefficient, the first concentration of the first element carbon, the first concentration of the second element carbon, the first concentration of the third element carbon, the second concentration of the second element carbon and the second concentration of the third element carbon, with the sum of the tenth optimal assimilation coefficient and the eleventh optimal assimilation coefficient being less than or equal to one as a constraint, the limited memory quasi-Newton optimization algorithm is used to fit the sixth objective function and the seventh objective function to determine the eleventh optimal assimilation coefficient and the twelfth optimal assimilation coefficient; The first optimal assimilation coefficient, the second optimal assimilation coefficient, the third optimal assimilation coefficient, the fourth optimal assimilation coefficient, the fifth optimal assimilation coefficient, the sixth optimal assimilation coefficient, the seventh optimal assimilation coefficient, the eighth optimal assimilation coefficient, the ninth optimal assimilation coefficient, the tenth optimal assimilation coefficient, the eleventh optimal assimilation coefficient and the twelfth optimal assimilation coefficient are determined as the optimal assimilation coefficient group.

3. The data assimilation method based on multiple linear regression model according to claim 1, characterized in that: The objective function group is: minδ1 2 =0.5*(a*OC1+b*OC2-OC1”) 2 ; minδ2 2 =0.5*[(1-a)*OC1+c*OC2+d*OC3-OC2”] 2 ; minδ3 2 =0.5*[(1-b-c)*OC2+e*OC3+f*OC4-OC3”] 2 ; minδ4 2 =0.5*[(1-d-e)*OC3+g*OC4+h*EC1-OC4”] 2 ; minδ5 2 =0.5*[(1-f-g)*OC4+i*EC1+j*EC2-EC1”] 2 ; minδ6 2 =0.5*[(1-h-i)*EC1+k*EC2+l*EC3-EC2”] 2 ; minδ7 2 =0.5*[(1-j-k)*EC2+(1-l)*EC3-EC3”] 2 ; Among them, δ1 is the relative standard deviation value between the assimilation data of the first organic carbon concentration and the second organic carbon concentration; δ2 is the relative standard deviation value between the assimilation data of the second organic carbon concentration and the second organic carbon concentration; δ3 is the relative standard deviation value between the assimilation data of the third organic carbon concentration and the second organic carbon concentration; δ4 is the relative standard deviation value between the assimilation data of the fourth organic carbon concentration and the second organic carbon concentration; δ5 is the relative standard deviation value between the assimilation data of the first element carbon concentration and the second element carbon concentration; δ6 is the relative standard deviation value between the assimilation data of the second element carbon concentration and the second element carbon concentration; δ7 is the relative standard deviation value between the assimilation data of the third element carbon concentration and the second element carbon concentration; a is the first optimal assimilation coefficient; b is the second optimal assimilation coefficient; c is the third optimal assimilation coefficient; d is the fourth optimal assimilation coefficient. coefficient; e is the fifth optimal assimilation coefficient; f is the sixth optimal assimilation coefficient; g is the seventh optimal assimilation coefficient; h is the eighth optimal assimilation coefficient; i is the ninth optimal assimilation coefficient; j is the tenth optimal assimilation coefficient; k is the eleventh optimal assimilation coefficient; l is the twelfth optimal assimilation coefficient; OC1 is the first concentration of the first organic carbon; OC2 is the first concentration of the second organic carbon; OC3 is the first concentration of the third organic carbon; OC4 is the first concentration of the fourth organic carbon; EC1 is the first concentration of the first element carbon; EC2 is the first concentration of the second element carbon; EC3 is the first concentration of the third element carbon; OC1” is the second concentration of the first organic carbon; OC2” is the second concentration of the second organic carbon; OC3” is the second concentration of the third organic carbon; OC4” is the second concentration of the fourth organic carbon; EC1” is the second concentration of the first element carbon; EC2” is the second concentration of the second element carbon; EC3” is the second concentration of the third element carbon.

4. The data assimilation method based on multiple linear regression model according to claim 1, characterized in that: The multiple linear regression model is: OC1' = a*OC1 + b*OC2; OC2'=(1-a)*OC1+c*OC2+d*OC3; OC3'=(1-bc)*OC2+e*OC3+f*OC4; OC4'=(1-de)*OC3+g*OC4+h*EC1; EC1'=(1-fg)*OC4+i*EC1+j*EC2; EC2'=(1-hi)*EC1+k*EC2+l*EC3; EC3'=(1-jk)*EC2+(1-l)*EC3; Among them, OC1' is the first organic carbon first concentration assimilation data; OC2' is the second organic carbon first concentration assimilation data; OC3' is the third organic carbon first concentration assimilation data; OC4' is the fourth organic carbon first concentration assimilation data; EC1' is the first element carbon first concentration assimilation data; EC2' is the second element carbon first concentration assimilation data; EC3' is the third element carbon first concentration assimilation data; OC1 is the first organic carbon first concentration; OC2 is the second organic carbon first concentration; OC3 is the third organic carbon first concentration; OC4 is the fourth The first concentration of organic carbon; EC1 is the first concentration of the first element carbon; EC2 is the first concentration of the second element carbon; EC3 is the first concentration of the third element carbon; a is the first optimal assimilation coefficient; b is the second optimal assimilation coefficient; c is the third optimal assimilation coefficient; d is the fourth optimal assimilation coefficient; e is the fifth optimal assimilation coefficient; f is the sixth optimal assimilation coefficient; g is the seventh optimal assimilation coefficient; h is the eighth optimal assimilation coefficient; i is the ninth optimal assimilation coefficient; j is the tenth optimal assimilation coefficient; k is the eleventh optimal assimilation coefficient; l is the twelfth optimal assimilation coefficient.

5. The data assimilation method based on multiple linear regression model according to claim 1, characterized in that: The DRI2001 carbon component analyzer uses a 633nm He / Ne laser and a FID detector to measure the first measurement concentration raw data; the DRI2015 carbon component analyzer uses a 635nm diode laser and a NDIR detector to measure the second measurement concentration raw data.

6. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the data assimilation method based on a multiple linear regression model as described in any one of claims 1 to 5.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the data assimilation method based on the multiple linear regression model described in any one of claims 1 to 5 is implemented.

8. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the data assimilation method based on the multiple linear regression model described in any one of claims 1 to 5 is implemented.

Images