Critical viscosity temperature prediction method and applications

Abstract

The present application relates to the field of coal gasification, and discloses a critical viscosity temperature prediction method and application, which comprises the following steps: (1) dividing N coal samples into calibration set samples and validation set samples, and obtaining characteristic spectral lines of the calibration set samples and the validation set samples by using laser-induced breakdown spectroscopy; (2) establishing a grey composition prediction model by using a partial least squares method through the characteristic spectral lines of the calibration set samples; (3) obtaining predicted values of all grey composition contents of the validation set samples through the grey composition prediction model; and (4) bringing the predicted values into an improved Urbain model to obtain a viscosity-temperature curve, and obtaining a critical viscosity temperature T CV through the viscosity-temperature curve; wherein N is greater than or equal to 44; and P2O5 is not considered in the grey composition of the improved Urbain model. The above method can greatly shorten the prediction time, the error fluctuation between the predicted value and the measured value of the critical viscosity temperature is smaller, and the prediction accuracy of the critical viscosity temperature of the coal sample is effectively improved.

Description

Technical Field

[0001] This invention relates to the field of coal gasification, specifically to a method and application for predicting critical viscosity temperature. Background Technology

[0002] Fluidized bed gasification technology with liquid slag discharge is currently the mainstream coal gasification technology. In this technology, the molten slag flow characteristics are one of the key factors for the long-term stable operation of the unit, affecting whether ash and slag can be continuously and smoothly discharged from the gasifier. Among these, molten slag viscosity, as a parameter measuring fluidity, is an essential characteristic of the melt structure. To ensure continuous and stable slag discharge, the furnace operating temperature should be higher than the critical viscosity temperature (Tcv). Therefore, obtaining the Tcv value quickly and accurately is crucial.

[0003] Traditional viscosity-temperature characteristic testing relies primarily on manual experiments, involving steps such as coal sample drying, grinding, screening, ash burning, ash fusion point testing (for reference), slag formation, crushed slag, high-temperature viscosity property testing, and data processing. This process is time-consuming and results in data lag, failing to meet the needs of rapid and accurate analysis of coal type adaptability in coal gasification processes. For a long time, numerous viscosity models have been developed, effectively predicting slag viscosity within certain composition and temperature ranges. For example, the modified Urbain model is used to predict the relationship between viscosity and temperature. These models are based on specific mathematical formulas, obtaining model parameters by fitting experimental data on coal sample ash composition. The ash composition data used is usually obtained through experimental testing (e.g., XRF testing), which also requires steps such as coal sample ash burning, resulting in time-consuming processes.

[0004] Laser-induced breakdown spectroscopy (LIBS) technology has many advantages, including speed, non-radioactivity, and ease of maintenance. Combined with algorithms, it can quickly obtain the ash composition of coal samples. For example, CN102053083A discloses an online measurement method for coal quality characteristics based on partial least squares (PLS). First, a LIBS system is used to detect calibrated coal samples with known compositions. Based on the obtained spectral intensity matrix, the calibrated samples are classified using PCA or PLS-DA methods. For each class of calibrated samples, a partial least squares (PLS) regression equation is established. Then, the sample to be tested is detected online to obtain the LIBS spectrum. The type of the sample to be tested is determined by PCA or PLS-DA methods. Finally, the spectral intensity matrix is ​​substituted into the corresponding type of PLS ​​regression equation to calculate the mass concentration of each element in the sample to be tested.

[0005] Starting from the practical problems of coal ash recycling and reuse, Wei Jiao et al. combined LIBS technology with chemometrics methods to construct different prediction models for quantitative analysis of coal ash composition and discrimination analysis of its types. First, the samples were screened for calibration and prediction sets, and anomalous spectra in the samples were removed using a combination of principal component analysis and Mahalanobis distance. The constructed PLS model was used to analyze seven main components in coal ash, obtaining good prediction results (correlation coefficient R > 0.9000).

[0006] Yao Shunchun et al. introduced partial least squares (PLS) as a quantitative analysis method for ash components in coal samples with multiple elements coexisting. The spectral data of 21 calibration samples were directly obtained by averaging three repeated measurements. Partial least squares was used as the quantitative analysis method for ash components. To verify the accuracy of the obtained PLS regression model in analyzing ash component content, six coal samples were used as test samples. The spectral data of these samples were substituted into the corresponding regression model for calculation to evaluate the accuracy of the quantitative analysis results.

[0007] However, the aforementioned existing technologies only provide prediction and analysis of ash composition, and do not address the T-value of coal samples. CV There is still a lack of accurate prediction methods. Summary of the Invention

[0008] The purpose of this invention is to overcome the problems of large errors and long time consumption in the prediction of critical viscosity temperature in the existing technology, and to provide a method and application for predicting critical viscosity temperature. This prediction method has a short test and calculation time and obtains a highly accurate critical viscosity temperature prediction value.

[0009] To achieve the above objectives, a first aspect of the present invention provides a method for predicting critical viscosity temperature, the method comprising:

[0010] (1) Divide N coal samples into a calibration set and a validation set, and use laser-induced breakdown spectroscopy to obtain the characteristic spectral lines of the calibration set and the validation set.

[0011] (2) Using the characteristic spectral lines of the calibration set samples obtained in step (1), a gray composition prediction model is established by partial least squares method;

[0012] (3) Using the gray composition prediction model, the predicted values ​​of all gray composition contents of the validation set samples are obtained;

[0013] (4) Substitute the predicted value into the improved Urbain model to obtain the viscosity-temperature curve, and obtain the critical viscosity temperature T through the viscosity-temperature curve. CV ;

[0014] Where N≥44; P2O5 is not considered in the gray composition of the improved Urbain model.

[0015] Preferably, in the ash composition of the improved Urbain model, Fe2O3 is considered as a neutral component.

[0016] The second aspect of this invention provides the application of the above-described critical viscosity temperature prediction method in coal gasification production.

[0017] Existing technologies typically use measured ash composition to predict critical viscosity temperature, which is time-consuming. This invention, however, proposes a novel method combining laser-induced breakdown spectroscopy (LIBS) testing with simulation to rapidly predict the critical viscosity temperature of coal samples. Through continuous research, the inventors developed an ash composition model based on LIBS testing and mathematical simulation to predict the ash composition of coal samples. This model is then combined with an improved modified Urbain model to obtain the viscosity-temperature relationship, ultimately leading to the critical viscosity temperature. This combination of LIBS testing, mathematical simulation, and the improved Urbain model enables rapid prediction. CV The prediction accuracy is high, and the required product quantity is small, which meets the needs of actual production and provides a reference for the production needs of coal gasification enterprises. Attached Figure Description

[0018] Figure 1 A schematic diagram of tangent processing for the viscosity-temperature curve of coal ash. Detailed Implementation

[0019] The endpoints and any values ​​of the ranges disclosed herein are not limited to the precise ranges or values, and these ranges or values ​​should be understood to include values ​​close to these ranges or values. For numerical ranges, the endpoint values ​​of the various ranges, the endpoint values ​​of the various ranges and individual point values, and individual point values ​​can be combined with each other to obtain one or more new numerical ranges, which should be considered as specifically disclosed herein.

[0020] The first aspect of this invention provides a method for predicting critical viscosity temperature, the method comprising:

[0021] (1) Divide N coal samples into a calibration set and a validation set, and use laser-induced breakdown spectroscopy to obtain the characteristic spectral lines of the calibration set and the validation set.

[0022] (2) Using the characteristic spectral lines of the calibration set samples obtained in step (1), a gray composition prediction model is established by partial least squares method;

[0023] (3) Using the gray composition prediction model, the predicted values ​​of all gray composition contents of the validation set samples are obtained;

[0024] (4) Substitute the predicted value into the improved Urbain model to obtain the viscosity-temperature curve, and obtain the critical viscosity temperature T through the viscosity-temperature curve. CV ;

[0025] Wherein, N≥44; in the ash composition of the improved Urbain model, P2O5 is not considered. In the prior art, there are numerous viscosity models, which can be mainly divided into structural models and empirical and semi-empirical models. The Urbain model classifies the oxides constituting the melt into three categories: acidic oxides, basic oxides, and neutral oxides. P2O5 and SiO2 are usually grouped together as acidic oxides. It is understood that the currently used Urbain model has also been modified. The modified Urbain model mentioned in this invention refers to the Urbain model used in the prior art. The improved Urbain model mentioned in this invention refers to a further improvement on the existing modified Urbain model. In this improved Urbain model, the influence of P2O5 is not considered in the ash composition, which can improve the accuracy of critical viscosity temperature prediction.

[0026] Furthermore, conventional methods for predicting critical viscosity temperature (CVT) obtain model parameters by fitting experimental data on coal sample ash composition. This ash composition data is typically obtained through experimental testing (e.g., XRF testing), but these methods are time-consuming. The inventors of this invention propose for the first time to obtain the predicted ash composition content using laser-induced breakdown spectroscopy combined with partial least squares method, and then combine this with an improved Urbain model for TVT prediction. CV The above prediction method can significantly shorten the prediction time while effectively improving the prediction accuracy of the critical viscosity temperature of coal samples.

[0027] According to the present invention, preferably, in the ash composition of the improved Urbain model, Fe2O3 is considered as a neutral component. Conventional Urbain models use normalized mole fractions of three types of oxides for calculations, such as binary systems (SiO2-M...). x O), ternary system (SiO2-Al2O3-M) x In viscosity prediction models for quaternary systems (SiO2-Al2O3-CaO-FeO), Fe2O3 is typically considered as FeO and is usually treated as an alkaline oxide. However, the inventors of this invention have discovered that by combining laser-induced breakdown spectroscopy with partial least squares method to obtain the predicted ash composition, and by treating Fe2O3 in the Urbain model as a neutral component, the accuracy of critical viscosity temperature prediction can be further improved.

[0028] According to the present invention, preferably, in the ash composition of the improved Urbain model, the acidic component is SiO2, the basic components include CaO, MgO, Na2O, K2O, TiO2, and MnO2, and the neutral components include Fe2O3 and Al2O3. By adopting the above preferred embodiment, further improving the calculation parameters of the Urbain model, and combining it with the ash composition content predicted by LIBS and partial least squares methods, the accuracy of predicting the critical viscosity temperature of coal samples can be effectively improved.

[0029] In this invention, the predicted value of the ash composition content in the sample to be tested is calculated by combining laser-induced breakdown spectroscopy and partial least squares method. It can adopt the simulation method known in the prior art. For example, you can refer to patent CN102053083A, the paper "Application Research of Laser-Induced Breakdown Spectroscopy Technology in Coal Industry" (Wei Jiao, Northwest University) and "Synchronous Measurement of Coal Ash Composition under Different Light Receiving Modes by LIBS Technology" (Yao Shunchun, Proceedings of the CSEE, 2013). The above patents or articles all provide specific calculation methods.

[0030] In this invention, the coal sample can be selected from any common coal type sample.

[0031] Preferably, step (1) includes: dividing the N coal samples into a calibration set and a validation set, and using laser-induced breakdown spectroscopy to measure the characteristic spectral lines of the calibration set and the validation set in the wavelength range of 180-800 nm.

[0032] In this invention, the validation set sample refers to the test sample composed of unknown gray.

[0033] According to the present invention, preferably, the method further includes vector normalization and mean centering of the obtained characteristic spectral lines. The vector normalization and mean centering can be performed using conventional methods in the art, well-known to those skilled in the art, as long as they can reduce spectral noise and increase the difference between sample spectra, thereby improving the robustness of model predictions.

[0034] According to the present invention, preferably, the gray composition prediction model is established by the following method:

[0035] (1-1) Obtain the absorbance matrix X and concentration vector Y of the graying elements by using the characteristic spectral lines of the calibration set samples;

[0036] (1-2) A correction model between the absorbance matrix X and the concentration vector Y is established using the partial least squares method, which is the gray composition prediction model.

[0037] According to the present invention, preferably, step (3) includes obtaining the absorbance matrix X of the graying elements through the characteristic spectral lines of the verification set samples.待测 The predicted values ​​of all gray composition contents of the validation set samples are obtained by substituting the gray composition prediction model into the model.

[0038] The specific calculation method can be found in Chapter 2 of the paper "Application Research of Laser-Induced Breakdown Spectroscopy in Coal Industry" (Wei Jiao, Northwest University), and will not be repeated here.

[0039] According to the present invention, the predicted value of the content of all ash components in the validation set sample is the mass fraction. Before being incorporated into the improved Urbain model, the mass fraction of each ash component is further converted into mole fraction × ash composition.

[0040] According to the present invention, the viscosity-temperature relationship of the improved Urbain model is well known to those skilled in the art, and it is sufficient as long as the composition of each component described above is satisfied. Preferably, the viscosity-temperature relationship in the improved Urbain model is as follows:

[0041] η=A×T×exp(1000B / T) / 10 (1)

[0042] lnA=-(0.29B+11.57) (2)

[0043] Where η is viscosity in Pa·s, A is pre-exponential factor, B is viscous activation energy in J / mol, and T is absolute temperature in K.

[0044] According to the present invention, preferably, the viscous activation energy B is calculated by the following formulas (3)-(17).

[0045] B = [X] CaO ×B CaO +X MgO ×B MgO ] / (X CaO +X MgO (3)

[0046] B CaO =B 0,CaO +B 1,CaO ×G+B 2,CaO ×G 2 +B 3,CaO ×G 3 (4)

[0047] B MgO =B 0,MgO +B 1,MgO ×G+B 2,MgO ×G 2 +B 3,MgO ×G 3 (5)

[0048] G=XSiO2 (6)

[0049] B 0,CaO =13.2+41.5a-45a 2 (7)

[0050] B 1,CaO =30.5 - 117.2a + 130a 2 (8)

[0051] B 2,CaO = -40.4 + 232.1a - 298.6a 2 (9)

[0052] B 3,CaO =60.8 - 156.4a + 213.6a 2 (10)

[0053] B 0,MgO =13.2 + 15.9a - 18.6a 2 (11)

[0054] B 1,MgO =30.5-54.1a+33a 2 (12)

[0055] B 2,MgO =-40.4+138a-112a 2 (13)

[0056] B 3,MgO =60.8 - 99.8a + 97.6a 2 (14)

[0057] a=M / (M+A') (15)

[0058] M = X CaO +X MgO +X Na2O +X K2O +X TiO2 +X MnO2 (16)

[0059] A' = X Al2O3 +X Fe2O3 (17)

[0060] Among them, X 灰组成 This is the predicted value for the corresponding ash composition content. For example, X Al2O3 This represents the mole fraction of Al2O3 in the sample.

[0061] In this invention, the predicted value of ash composition content is substituted into the above formulas (1)-(17) to obtain the viscosity-temperature curve. Then, the measured viscosity-temperature curve of coal ash is processed by tangent processing to obtain the coordinates of the intersection of the two tangents. The temperature corresponding to the intersection coordinates is the critical viscosity temperature T. CV The processing procedure can be referenced. Figure 1 The above processing procedures are well known to those skilled in the art and will not be described in detail here.

[0062] The second aspect of this invention provides the application of the above-described critical viscosity temperature prediction method in coal gasification production.

[0063] The prediction method described in this invention uses a small amount of product, can achieve rapid prediction, has high accuracy, meets the needs of actual production, and provides a reference for the production needs of coal gasification enterprises.

[0064] The present invention will be described in detail below through embodiments.

[0065] The T of coal samples in the following examples CV，实测 Measured according to the method of DL / T 660-2007.

[0066] The coal samples used in the following examples were from China Petroleum & Chemical Corporation Zhenhai Refining & Chemical Branch and China Petrochemical Corporation Maoming Petrochemical Co., Ltd.

[0067] Example 1

[0068] (1) A total of 100 coal samples were collected. The characteristic spectral lines of each coal sample in the range of 180-800 nm were measured by LIBS. The characteristic spectral lines were then processed by vector normalization and mean centering.

[0069] (2) The coal samples were divided into a calibration set and a validation set, with 84 calibration set samples and 16 validation set samples.

[0070] Following the method disclosed in Chapter 2 of the paper "Application Research of Laser-Induced Breakdown Spectroscopy in Coal Industry" (Wei Jiao, Northwest University), the absorbance matrix X and concentration vector Y of ash-forming elements are obtained by correcting the characteristic spectral lines of the sample set. A partial least squares (PLS) prediction model for ash composition is then established. The absorbance matrix X of ash-forming elements is obtained by validating the characteristic spectral lines of the sample set. 待测 The corresponding ash composition is obtained by inputting it into the ash composition prediction model, including SiO2, Al2O3, Fe2O3, CaO, MgO, Na2O, K2O, TiO2, MnO2, and SO3.

[0071] (3) The predicted ash composition is expressed as a weight percentage, which is then converted into a mole fraction (X). 灰组成Then, the relevant parameters of the improved Urbain model are calculated according to the formula to obtain the viscosity-temperature curve. Tangents are processed on the measured viscosity-temperature curve to obtain the coordinates of the intersection point of the two tangents. The temperature corresponding to the intersection point is the critical viscosity temperature T. CV预测 The results are shown in Table 1.

[0072] G=X SiO2

[0073] M = X CaO +X MgO +X Na2O +X K2O +X TiO2 +X MnO2

[0074] A' = X Al2O3 +X Fe2O3

[0075] a = M / (M + A')

[0076] B 0,CaO =13.2+41.5a-45a 2

[0077] B 1,CaO =30.5 - 117.2a + 130a 2

[0078] B 2,CaO = -40.4 + 232.1a - 298.6a 2

[0079] B 3,CaO =60.8 - 156.4a + 213.6a 2

[0080] B CaO =B 0,CaO +B 1,CaO ×G+B 2,CaO ×G 2 +B 3,CaO ×G 3

[0081] B 0,MgO =13.2 + 15.9a - 18.6a 2

[0082] B 1,MgO =30.5-54.1a+33a 2

[0083] B 2,MgO =-40.4+138a-112a 2

[0084] B 3,MgO =60.8 - 99.8a + 97.6a 2

[0085] B MgO =B 0,MgO +B 1,MgO ×G+B 2,MgO ×G 2 +B 3,MgO ×G 3

[0086] B = [X] CaO ×B CaO +X MgO ×B MgO ] / (X CaO +X MgO )

[0087] η=A×T×exp(1000B / T) / 10

[0088] lnA = -(0.29B + 11.57)

[0089] T calculated from each validation set sample CV，预测 Compared with the T measured in the laboratory CV，实测 In comparison, when the absolute value of the difference is less than or equal to 50°C, the probability is 75%.

[0090] Table 1

[0091]

[0092]

[0093] Comparative Example 1

[0094] The ash composition of the validation set samples in Example 1 was determined using XRF, including SiO2, Al2O3, Fe2O3, CaO, MgO, Na2O, K2O, TiO2, MnO2, SO3, and P2O5. The weight fractions of the ash composition were converted to mole fractions to calculate the model parameters of the modified Urbain model. In the parameters, acidic components include SiO2 and P2O5; basic components include CaO, MgO, Na2O, K2O, TiO2, and MnO2; Fe2O3 and Al2O3 are considered as neutral components. The calculation method for the modified Urbain model parameters is as follows:

[0095] G'=X' SiO2 +X' P2O5

[0096] M'=X' CaO +X' MgO +X' Na2O +X'K2O +X’ TiO2 +X’ MnO2

[0097] A1’=X’ Al2O3 +X’ Fe2O3

[0098] a’=M’ / (M’+A1’)

[0099] B’ 0,CaO =13.2+41.5a’-45(a’) 2

[0100] B’ 1,CaO =30.5-117.2a’+130(a’) 2

[0101] B’ 2,CaO =-40.4+232.1a’-298.6(a’) 2

[0102] B’ 3,CaO =60.8-156.4a’+213.6(a’) 2

[0103] B’ CaO =B’ 0,CaO +B’ 1,CaO ×G’+B’ 2,CaO ×(G’) 2 +B’ 3,CaO ×(G’) 3

[0104] B’ 0,MgO =13.2+15.9a’-18.6(a’) 2

[0105] B’ 1,MgO =30.5-54.1a’+33(a’) 2

[0106] B’ 2,MgO =-40.4+138a’-112(a’) 2

[0107] B’ 3,MgO =60.8-99.8a’+97.6(a’) 2

[0108] B’ MgO =B’ 0,MgO +B’ 1,MgO ×G’+B’ 2,MgO ×(G’) 2 +B’3,MgO ×(G') 3

[0109] B' = [X' CaO ×B' CaO +X' MgO ×B' MgO ] / (X' CaO +X' MgO )

[0110] η'=A'×T'×exp(1000B' / T') / 10

[0111] lnA' = -(0.29B' + 11.57)

[0112] Where X' 灰组成 The mole fraction of ash composition is used. The viscosity-temperature curve is obtained through model parameters. Tangents are then applied to the curve, and the coordinates of the intersection point of the two tangents are determined. The temperature corresponding to the intersection point is T. CV2，预测 The results are shown in Table 2. CV2，预测 Compared with the T measured in the laboratory CV，实测 In comparison, when the absolute value of the difference is less than or equal to 50℃, the probability is 56%.

[0113] Table 2

[0114]

[0115]

[0116] Comparative Example 2

[0117] The coal sample ash composition was obtained using the same steps as in Example 1, and used to calculate the model parameters of the modified Urbain model. The acidic components in the parameters include SiO2 and P2O5; the alkaline components include CaO, MgO, Na2O, K2O, TiO2, and MnO2; and the neutral component is Al2O3. The calculation method for the modified Urbain model parameters is as follows:

[0118] G'=X' SiO2 +X' P2O5

[0119] M'=X' CaO +X' MgO +X' Na2O +X' K2O +X' TiO2 +X' MnO2

[0120] A2'=X' Al2O3

[0121] a' = M' / (M' + A2')

[0122] B’ 0,CaO =13.2+41.5a’-45(a’) 2

[0123] B’ 1,CaO =30.5-117.2a’+130(a’) 2

[0124] B’ 2,CaO =-40.4+232.1a’-298.6(a’) 2

[0125] B’ 3,CaO =60.8-156.4a’+213.6(a’) 2

[0126] B’ CaO =B’ 0,CaO +B’ 1,CaO ×G’+B’ 2,CaO ×(G’) 2 +B’ 3,CaO ×(G’) 3

[0127] B’ 0,MgO =13.2+15.9a’-18.6(a’) 2

[0128] B’ 1,MgO =30.5-54.1a’+33(a’) 2

[0129] B’ 2,MgO =-40.4+138a’-112(a’) 2

[0130] B’ 3,MgO =60.8-99.8a’+97.6(a’) 2

[0131] B’ MgO =B’ 0,MgO +B’ 1,MgO ×G’+B’ 2,MgO ×(G’) 2 +B’ 3,MgO ×(G’) 3

[0132] B’=[X’ CaO ×B’ CaO +X’ MgO ×B’ MgO ] / (X’ CaO +X’MgO )

[0133] η'=A'×T'×exp(1000B' / T') / 10

[0134] lnA' = -(0.29B' + 11.57)

[0135] Where X' 灰组成 The mole fraction of ash composition is used. The viscosity-temperature curve is obtained through model parameters. Tangents are then applied to the curve, and the coordinates of the intersection point of the two tangents are determined. The temperature corresponding to the intersection point is T. CV3，预测 The results are shown in Table 3. CV3，预测 Compared with the T measured in the laboratory CV，实测 In comparison, when the absolute value of the difference is less than or equal to 50°C, the probability is 50%.

[0136] Table 3

[0137] 1 1120 1158 -38 2 1125 1058 67 3 1185 1220 -35 4 1195 1166 29 5 1341 1191 150 6 1325 1238 87 7 1107 1208 -101 8 1230 1220 10 9 1172 1248 -76 10 1135 1220 -85 11 1176 1205 -29 12 1182 1209 -27 13 1258 1191 67 14 1325 1270 55 15 1255 1212 43 16 1185 1150 35

[0138] As can be seen from the results in Tables 1-3, the present invention obtains the predicted ash composition content by combining laser-induced breakdown spectroscopy with partial least squares method, and then combines it with the improved Urbain model. The above prediction method can significantly shorten the prediction time while reducing the error fluctuation between the predicted and measured values ​​of critical viscosity temperature, effectively improving the prediction accuracy of critical viscosity temperature of coal samples.

[0139] The preferred embodiments of the present invention have been described in detail above; however, the present invention is not limited thereto. Within the scope of the inventive concept, various simple modifications can be made to the technical solutions of the present invention, including combinations of various technical features in any other suitable manner. These simple modifications and combinations should also be considered as the content disclosed in the present invention and are all within the protection scope of the present invention.

Claims

1. A method for predicting critical viscosity temperature, characterized in that, The method includes: (1) Divide N coal samples into a calibration set and a validation set, and use laser-induced breakdown spectroscopy to obtain the characteristic spectral lines of the calibration set and the validation set. (2) Using the characteristic spectral lines of the calibration set samples obtained in step (1), a gray composition prediction model is established using the partial least squares method; The gray composition prediction model is established using the following method: (1-1) Obtain the absorbance matrix X and concentration vector Y of the graying elements by using the characteristic spectral lines of the calibration set samples; (1-2) A calibration model between the absorbance matrix X and the concentration vector Y is established using the partial least squares method, which is the gray composition prediction model; (3) Using the gray composition prediction model, the predicted values ​​of all gray composition contents of the validation set samples are obtained; (4) Substitute the predicted value into the improved Urbain model to obtain the viscosity-temperature curve, and obtain the critical viscosity temperature T through the viscosity-temperature curve. CV ; Wherein, N≥44; in the gray composition of the improved Urbain model, P2O5 is not considered; In the ash composition of the improved Urbain model, the acidic component is SiO2, the alkaline components include CaO, MgO, Na2O, K2O, TiO2 and MnO2, and the neutral components include Fe2O3 and Al2O3. The viscosity-temperature relationship in the improved Urbain model is as follows: =A×T×exp(1000B / T) / 10(1) lnA = - (0.29B + 11.57) (2) in, Where A is viscosity (Pa·s), B is pre-exponential factor (A), and T is viscosity activation energy (J / mol). The B is calculated using the following formulas (3)-(17). B=[X CaO ×B CaO + X MgO ×B MgO ] / ( X CaO +X MgO )(3) B CaO = B 0,CaO + B 1,CaO ×G+ B 2,CaO ×G 2 + B 3,CaO ×G 3 (4) B MgO = B 0,MgO + B 1,MgO ×G+ B 2,MgO ×G 2 + B 3,MgO ×G 3 (5) G=X SiO2 (6) B 0,CaO =13.2 + 41.5a - 45a 2 (7) B 1,CaO =30.5 - 117.2a + 130a 2 (8) B 2,CaO =-40.4+232.1a-298.6a 2 (9) B 3,CaO =60.8-156.4a+213.6a 2 (10) B 0,MgO =13.2 + 15.9a - 18.6a 2 (11) B 1,MgO =30.5 - 54.1a + 33a 2 (12) B 2,MgO =-40.4+138a-112a 2 (13) B 3,MgO =60.8-99.8a+97.6a 2 (14) a = M / (M + A') (15) M=X CaO +X MgO +X Na2O +X K2O +X TiO2 +X MnO2 (16) A’=X Al2O3 +X Fe2O3 (17) Among them, X 灰组成 This represents the predicted value for the corresponding ash composition content.

2. The prediction method according to claim 1, wherein, N=44-100。 3. The prediction method according to claim 1 or 2, wherein, Step (1) includes: using laser-induced breakdown spectroscopy to measure the characteristic spectral lines of the calibration set sample and the validation set sample in the wavelength range of 180-800 nm.

4. The prediction method according to claim 1, wherein, The ratio of the number of samples in the calibration set to the number of samples in the validation set is ≥1.5:

1.

5. The prediction method according to claim 1, wherein, The method further includes vector normalization and mean centering of the obtained feature spectral lines.

6. The prediction method according to claim 1, wherein, Step (3) includes obtaining the absorbance matrix X of the graying elements through the characteristic spectral lines of the validation set samples. 待测 The predicted values ​​of all gray composition contents of the validation set samples are obtained by substituting the gray composition prediction model into the model.

7. The application of the critical viscosity temperature prediction method according to any one of claims 1-6 in coal gasification production.

Images