A closed flash point real-time prediction method based on micro distillation range measuring device

Abstract

The application provides a closed flash point real-time prediction method based on a micro distillation range determination device, relates to the technical field of distillation range and flash point testing, and comprises the following steps: acquiring a pre-sequenced distillation range temperature group in real time, the pre-sequenced distillation range temperature group comprising initial vapor temperatures corresponding to a plurality of preset recovery volume percentage nodes, acquiring a vapor pressure and a liquid temperature in real time to obtain a vapor pressure change rate and a liquid temperature change gradient, and then calculating a heat capacity disturbance factor of the device; using the heat capacity disturbance factor to correct each initial vapor temperature in the pre-sequenced distillation range temperature group in real time to obtain corresponding first corrected vapor temperatures; based on a weight distribution relationship of a linear correlation characteristic between a test vapor temperature of the preset recovery volume percentage node and the closed flash point, the closed flash point is predicted through linear weighting calculation of each first corrected vapor temperature. The application realizes the closed flash point prediction while the distillation range is determined, and time and samples are significantly saved.

Description

Technical Field

[0001] This application relates to the field of distillation range and flash point testing technology, specifically to a method for real-time prediction of closed-cup flash point based on a micro-distillation range measuring device. Background Technology

[0002] Distillation range and closed-cup flash point are two extremely important basic physicochemical indicators for evaluating the evaporation performance, safety, and performance in use of petroleum products. Distillation range reflects the distillation characteristics of petroleum products at different temperatures, while closed-cup flash point directly characterizes the temperature at which the flammable vapors released by the petroleum product reach the minimum combustion concentration under specific conditions. It is a key parameter for the safe storage, transportation, and use of petroleum products.

[0003] In existing technologies, distillation range determination (such as standard methods like ASTM D86 and GB / T 6536) and closed-cup flash point determination (such as standard methods like ASTM D93 and GB / T 261) are two completely independent tests. Each test requires separate sampling, consuming more sample volume, which is particularly disadvantageous for trace or precious samples. Furthermore, it incurs additional time costs, as completing each test requires separate instrument preparation, sample loading, test execution, and result analysis, significantly increasing the overall testing time. Summary of the Invention

[0004] This application provides a method for real-time prediction of closed-cup flash point based on a micro-distillation range measuring device, which solves the technical problems of existing distillation range measuring methods being unable to predict closed-cup flash point simultaneously, having high cost, and long testing time. This application is achieved through the following solution: The preceding distillation range temperature group is acquired in real time. The preceding distillation range temperature group includes the initial vapor temperature corresponding to multiple preset recovery volume percentage nodes. The preset recovery volume percentage nodes are located in the initial rising segment of the boiling range curve of the micro-distillation range measuring device, and the linear correlation coefficient with the closed-cup flash point is greater than or equal to a first preset threshold. Real-time acquisition of steam pressure and liquid temperature is used to obtain the rate of change of steam pressure and the gradient of change of liquid temperature, and then the thermal capacity disturbance factor of the device is calculated. The heat capacity perturbation factor is used to correct the initial steam temperature of each of the preceding distillation temperature groups in real time to obtain the corresponding first corrected steam temperature. Based on the pre-established weighting relationship of the linear correlation between the test steam temperature and the closed-cup flash point of the preset recovery volume percentage node, the closed-cup flash point is predicted by linear weighting of each of the first corrected steam temperatures.

[0005] Furthermore, the real-time acquisition of the preceding distillation range temperature group includes the following steps: Determine the fuel type of the fuel to be tested; Based on the fuel type, a corresponding preset node set is obtained, and the preset node set includes multiple preset recycling volume percentage nodes; During the micro-distillation range determination process, when the recovery volume percentage reaches each preset recovery volume percentage node, the initial steam temperature measured by the steam temperature detection unit is recorded in real time to form a preceding distillation range temperature group.

[0006] Furthermore, obtaining the corresponding preset node set based on the fuel type includes the following steps: The preset node database is retrieved and traversed to obtain a preset node set corresponding to the fuel type of the fuel to be tested; the preset node database includes multiple preset fuel types and preset node sets corresponding to each fuel type; The method also includes constructing the preset node database; The construction of the preset node database includes the following steps: Multiple groups of fuel samples were selected, and the distillation range test of the standard distillation range test method was performed on each group of fuel samples. The test vapor temperature of all test nodes between the initial boiling point and 10% recovery volume percentage was recorded. Simultaneously perform closed-cup flash point testing using the standard closed-cup flash point test method on each group of fuel samples to obtain the test closed-cup flash point; For each group of fuel samples, calculate the linear correlation coefficient between the test vapor temperature and the test closed-cup flash point for each test node of the fuel sample in that group; The test nodes whose linear correlation coefficient is greater than or equal to the first preset threshold are retained as preset recovery volume percentage nodes corresponding to the group of fuel samples, thereby obtaining the preset node set corresponding to the group of fuel samples; Wherein, the first preset threshold is greater than or equal to 0.99.

[0007] Furthermore, the fuel type includes jet fuel or diesel fuel; The step of obtaining the corresponding preset node set based on the fuel type includes the following steps: If the fuel type is jet fuel or diesel, the corresponding preset node set includes four preset recovery volume percentage nodes of 1%, 2%, 3% and 4%.

[0008] Furthermore, the preset node database also includes preset initial boiling points corresponding to each group of preset fuel types; The determination of the fuel type of the fuel to be tested includes the following steps: The initial boiling point of the fuel to be tested is obtained during the micro-boiling range determination process; Retrieve and traverse the preset node database to determine whether there is a preset initial boiling point in the preset node database that matches the measured initial boiling point; If a preset initial boiling point exists that matches the measured initial boiling point, then the preset fuel type corresponding to the preset initial boiling point that matches the measured initial boiling point is taken as the fuel type of the fuel to be tested.

[0009] According to the technical solution provided in this application, the preset node database also includes standard gradient template vectors corresponding to each group of preset fuel types; After determining whether a preset initial boiling point matching the measured initial boiling point exists in the preset node database, the method further includes the following steps: If there is no preset initial boiling point that matches the measured initial boiling point, the steam temperature detection values ​​of at least three equally spaced nodes of the fuel to be tested are extracted in real time within the range from the initial boiling point to 5% recovery volume percentage to form a characteristic temperature sequence. Calculate the first-order difference gradient vector of the characteristic temperature sequence; Traverse the standard gradient template vectors corresponding to each group of the preset fuel types in the preset node database, and calculate the similarity between the first-order difference gradient vector and each of the standard gradient template vectors respectively; The preset fuel type corresponding to the standard gradient template vector that minimizes the similarity is selected as the fuel type of the fuel to be tested.

[0010] Furthermore, after obtaining the corresponding first corrected steam temperature, the process further includes the following steps: Real-time acquisition of ambient atmospheric pressure; The method of predicting the closed-cup flash point by linear weighting of each of the first corrected steam temperatures includes the following steps: If the deviation between the ambient atmospheric pressure and the standard atmospheric pressure is less than or equal to the second preset threshold, the closed-cup flash point is predicted by linear weighted calculation of each of the first corrected vapor temperatures. After obtaining the real-time ambient atmospheric pressure, the following steps are also included: If the deviation between the ambient atmospheric pressure and the standard atmospheric pressure is greater than the second preset threshold, the first corrected steam temperature is corrected based on the deviation between the ambient atmospheric pressure and the standard atmospheric pressure to obtain the second corrected steam temperature. The closed-cup flash point is predicted by linear weighting of each of the second corrected steam temperatures.

[0011] Furthermore, after acquiring the preceding distillation range temperature set in real time, the method further includes the following steps: The sliding standard deviation of the rate of change of steam temperature for adjacent preset recovery volume percentage nodes in the preceding distillation temperature group is calculated in real time. The step of using the heat capacity perturbation factor to correct the initial steam temperature in each of the preceding distillation range temperature groups in real time to obtain the corresponding first corrected steam temperature includes the following steps: If the rate of change of steam temperature at all the preset recovery volume percentage nodes does not deviate from the current moving average by more than 3 times the moving standard deviation, then the heat capacity disturbance factor is used to correct the initial steam temperature in the preceding distillation temperature group in real time to obtain the corresponding first corrected steam temperature.

[0012] Furthermore, after calculating the sliding standard deviation of the steam temperature change rate of adjacent preset recovery volume percentage nodes in the preceding distillation temperature group in real time, the method further includes the following steps: If the steam temperature change rate of the preset recovery volume percentage node deviates from the current moving average by more than 3 times the moving standard deviation, then the preset recovery volume percentage node whose steam temperature change rate deviates from the current moving average by more than 3 times the moving standard deviation is marked as an abnormal node. The initial steam temperature corresponding to the abnormal node is corrected by linear interpolation of the initial steam temperature of the two preset recovery volume percentage nodes adjacent to the abnormal node.

[0013] Furthermore, after calculating the sliding standard deviation of the steam temperature change rate of adjacent preset recovery volume percentage nodes in the preceding distillation temperature group in real time, the method further includes the following steps: If two or more of the preset recovery volume percentage nodes are marked as abnormal nodes, the liquid temperature detection values ​​at the time of the abnormal node corresponding to all the abnormal nodes are extracted in real time. Based on the liquid temperature detection value at the time of the abnormal node and the temperature offset of the heat capacity disturbance factor, the compensation steam temperature corresponding to each abnormal node is obtained. The initial steam temperature is corrected using the compensation steam temperature corresponding to each of the aforementioned abnormal nodes.

[0014] Compared with the prior art, the beneficial effects of this application are as follows: I. Highly Integrated Testing Process, Doubly Increased Efficiency: This application achieves simultaneous prediction of the closed-cup flash point during a single micro-range distillation test. This application eliminates the need to interrupt the distillation range test or use additional flash point measuring equipment; it directly uses the real-time data collected by the distillation range meter itself to complete the flash point prediction. This fundamentally eliminates the drawback of traditional methods requiring two independent tests. II. Significant Time and Sample Savings: Since flash point prediction is performed simultaneously during the distillation range test, no additional testing time is required. Furthermore, only the sample volume required for a single trace distillation range test is needed to simultaneously obtain distillation range data and flash point prediction values, greatly saving valuable testing time and sample resources. This is particularly suitable for high-throughput screening, R&D stages, or situations with limited sample quantities. III. Real-time Dynamic Prediction and Rapid Response: This application achieves true real-time prediction. In the initial stage of the distillation range test (the rising section of the boiling range curve), the system begins collecting key node temperatures, vapor pressures, and liquid temperatures, calculating the heat capacity disturbance factor in real time and performing temperature corrections, ultimately outputting the flash point prediction value through weighted calculation. This allows users to obtain flash point information before the end of the distillation range test, accelerating the decision-making process. IV. Improved Prediction Accuracy: This application ensures a stronger intrinsic correlation between the base temperature data used for prediction and the flash point by selecting a preset recovery volume percentage node (located in the initial rising segment of the boiling range curve) with a high linear correlation coefficient with the closed-cup flash point. A heat capacity disturbance factor is introduced, and calculations are performed using real-time acquired vapor pressure change rate and liquid temperature change gradient to dynamically correct the impact of thermal fluctuations in the test system (such as uneven heating and pressure changes) on the initial vapor temperature measurement. This real-time temperature correction significantly improves the accuracy of the base data used for prediction. Based on a weighting relationship established using historical data or theoretical analysis, higher weights are assigned to corrected temperature nodes with stronger correlation to the flash point, making the final weighted calculation of the predicted closed-cup flash point more reliable. Attached Figure Description

[0015] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein: Figure 1 This is a flowchart illustrating the steps of the real-time prediction method for closed-cup flash point based on a micro-distillation range measuring device provided in this application embodiment. Detailed Implementation

[0017] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0018] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0019] Example 1 As mentioned in the background section, to address the problems in the prior art, this application proposes a real-time prediction method for the closed-cup flash point based on a micro-distillation range measuring device. This method is implemented using a micro-distillation range measuring device, which includes an integrated sensing component connected to the outlet end of the distillation flask via a sealed interface. The integrated sensing component includes a liquid temperature detection section extending below the liquid surface of the liquid to be distilled; a vapor temperature detection section suspended in the vapor region within the distillation chamber; and a vapor pressure detection section for detecting the pressure within the distillation chamber. like Figure 1 As shown, the method includes the following steps: S1. Real-time acquisition of the preceding distillation range temperature group, the preceding distillation range temperature group includes the initial steam temperature corresponding to multiple preset recovery volume percentage nodes, the preset recovery volume percentage nodes are located in the initial rising segment of the boiling range curve of the micro-distillation range measuring device, and the linear correlation coefficient with the closed-cup flash point is greater than or equal to the first preset threshold; wherein, the initial rising segment is the local curve segment corresponding to the boiling range curve from the initial boiling point to 10% recovery volume.

[0020] Furthermore, the real-time acquisition of the preceding distillation range temperature group includes the following steps: Determine the fuel type of the fuel to be tested; Based on the fuel type, a corresponding preset node set is obtained, and the preset node set includes multiple preset recycling volume percentage nodes; During the micro-distillation range determination process, when the recovery volume percentage reaches each preset recovery volume percentage node, the initial steam temperature measured by the steam temperature detection unit is recorded in real time to form a preceding distillation range temperature group.

[0021] S2. Real-time acquisition of steam pressure and liquid temperature to obtain the steam pressure change rate and liquid temperature change gradient, and then calculation of the thermal capacity disturbance factor of the device. Specifically, the heat capacity disturbance factor reflects the delay effect of the system's thermal inertia on steam temperature measurement; the steam pressure change rate (ΔP / Δt) represents the pressure change per unit time; and the liquid temperature change gradient (ΔT / Δt) represents the rate of liquid temperature rise per unit time, expressed by the formula... The heat capacity perturbation factor is obtained, where HDF represents the heat capacity perturbation factor, and k1 and k2 are weighting coefficients determined through calibration experiments, for example, k1=0.35 and k2=0.65.

[0022] S3. Use the heat capacity disturbance factor to correct the initial steam temperature of each of the preceding distillation temperature groups in real time to obtain the corresponding first corrected steam temperature. Specifically, using HDF via the formula Dynamic compensation is performed on the initial steam temperature, where T 修正 T represents the first corrected steam temperature. 初始 The initial steam temperature is represented by α, which is a temperature correction factor obtained through thermodynamic calibration.

[0023] S4. Based on the pre-established weighted distribution relationship between the test steam temperature and the closed-cup flash point of the preset recovery volume percentage node, the closed-cup flash point is predicted by linear weighted calculation of each of the first corrected steam temperatures. The initial steam temperature is obtained by the steam temperature detection unit, the steam pressure is obtained by the steam pressure detection unit, and the liquid temperature is obtained by the liquid temperature detection unit.

[0024] Specifically, the preceding distillation range temperature set is obtained by real-time acquisition of the initial steam temperatures corresponding to multiple preset recovery volume percentage nodes during the initial stage of distillation range testing (the rising segment of the boiling range curve). The preceding distillation range temperature set refers to the set of node temperatures in the initial rising segment of the boiling range curve (usually from the initial boiling point to 10% recovery volume) that are strongly correlated with the closed-cup flash point (linear correlation coefficient ≥ 0.99). The preset recovery volume percentage nodes are selected through verification using historical data; for example, nodes from 1% to 4% are selected for jet fuel. The pre-established weighted relationship between the test steam temperature and the closed-cup flash point of the preset recovery volume percentage nodes can also be pre-stored in the preset node database for easy retrieval and use.

[0025] Furthermore, obtaining the corresponding preset node set based on the fuel type includes the following steps: The preset node database is retrieved and traversed to obtain a preset node set corresponding to the fuel type of the fuel to be tested; the preset node database includes multiple preset fuel types and preset node sets corresponding to each fuel type; The method also includes constructing the preset node database; The construction of the preset node database includes the following steps: Multiple groups of fuel samples were selected, and the distillation range test of the standard distillation range test method was performed on each group of fuel samples. The test vapor temperature of all test nodes between the initial boiling point and 10% recovery volume percentage was recorded. Simultaneously perform closed-cup flash point testing using the standard closed-cup flash point test method on each group of fuel samples to obtain the test closed-cup flash point; For each group of fuel samples, calculate the linear correlation coefficient between the test vapor temperature and the test closed-cup flash point for each test node of the fuel sample in that group; The test nodes whose linear correlation coefficient is greater than or equal to the first preset threshold are retained as preset recovery volume percentage nodes corresponding to the group of fuel samples, thereby obtaining the preset node set corresponding to the group of fuel samples; Wherein, the first preset threshold is greater than or equal to 0.99.

[0026] Furthermore, the fuel type includes jet fuel or diesel fuel; The step of obtaining the corresponding preset node set based on the fuel type includes the following steps: If the fuel type is jet fuel or diesel, the corresponding preset node set includes four preset recovery volume percentage nodes of 1%, 2%, 3% and 4%.

[0027] The following describes the pre-establishment process of the preset node database: The scientific basis is that the light fraction temperature (1%~4% recovery volume) dominates the flash point of oil products, as it directly reflects the concentration of volatile components. The weights essentially represent the contribution of each temperature node to flash point prediction, determined by the strength of the linear correlation between that node temperature and the closed-cup flash point. Example: If the 2% temperature (T2) has the strongest correlation with the flash point (R... 2 If the value is 0.995, then assign it the highest weight (e.g., 0.198); establish the flash point prediction equation through multiple linear regression: Where CFP represents the closed-cup flash point, T1, T2, ..., T n This indicates the test steam temperature at the test node corresponding to n% of the recovered volume. , … The weight corresponding to each test steam temperature is represented by C, which is a constant term representing the intercept. The following tests were conducted using jet fuel or diesel as fuel samples. The specific test procedure is as follows: Sample selection: covering the entire range of target fuels (e.g., 10 jet fuels + 10 diesels); Distillation range test: executing the standard test methods of GB / T6536 / NB SH / T 6071, recording the temperatures of all nodes from the initial boiling point to 10% (1%, 2%, ..., 10%); Flash point test simultaneously executing GB / T 261 to obtain the baseline closed-cup flash point value. For each recovery node (e.g., 1%, 2%, ..., 10%), the Pearson correlation coefficient between its temperature and closed-cup flash point is calculated. Taking jet fuel as an example: 1% temperature: r1 = 0.992; 2% temperature: r2 = 0.995, corresponding to the optimal correlation; 4% temperature: r4 = 0.991; 10% temperature: r 10 =0.972, remove (R ²<0.99), Input variable: filter r i For nodes with a value ≥0.99 (1%, 2%, 3%, 4%), the regression algorithm uses least squares to fit the weights. With the constant C (representing the reference flash point temperature for this fuel type), the fitting result for jet fuel or diesel is: CFP = 0.193T1 + 0.198T2 + 0.193T3 + 0.183T4 − 77.9, meaning the preset recovery volume percentage nodes for this group of fuel types (jet fuel) are 1%, 2%, 3%, and 4%. The pre-established weighted relationship between the linear correlation characteristics of the test steam temperature and closed-cup flash point for the preset recovery volume percentage nodes is CFP = 0.193T1 + 0.198T2 + 0.193T3 + 0.183T4 − 77.9, where the weight allocation is based on: =0.198 (maximum weight) corresponds to the strongest 2% temperature correlation. =0.183 (minimum weight), corresponding to a slightly weaker 4% temperature correlation. When using this method to predict the flash point of fuel, the obtained preceding distillation temperature set is substituted into CFP=0.193T1+0.198T2+0.193T3+0.183T4−77.9 to obtain the predicted closed-cup flash point.

[0028] Specifically, the test results of 10 samples using both GB / T 6536 and NB / SH / T 6071 methods were substituted into the calculation according to the formula CFP=0.193T1+0.198T2+0.193T3+0.183T4−77.9, and compared with the test results of GB / T 261 method, as shown in Table 1: Table 1 ; As shown in Table 1, the maximum range between the calculated closed-cup flash point of the 10 samples and the test results of the GB / T 261 method is only 4.1℃, and the reproducibility R=4.1 / 92.5=0.044, which is better than the reproducibility requirement of GB / T 261 method (0.071).

[0029] In addition, three No. 3 jet fuel samples, three diesel samples, and two closed-cup flash point standards were selected. The closed-cup flash point was determined using the GB / T 261 method, and the distillation range was determined using the NB / SH / T 6071 method. Based on the distillation range results, the closed-cup flash point was calculated using the formula CFP=0.193T1+0.198T2+0.193T3+0.183T4−77.9. The results are compared in Table 2. As can be seen from Table 2, the maximum deviation of the calculated closed-cup flash point for the eight samples was 0.6℃, and the repeatability r=0.6 / 45.3=0.013, which is better than the repeatability requirement of the GB / T 261 method (0.029X). The maximum range of the test results compared to the GB / T 261 method is only 2.6℃, and the reproducibility R = 2.6 / 93 = 0.028, which is better than the reproducibility requirement of the GB / T 261 method (0.071X). It should be noted that this method can be applied to various fuel types, including but not limited to jet fuel and diesel; Table 2 shows: Table 2 ; In a preferred embodiment, the preset node database further includes preset initial boiling points corresponding to each group of preset fuel types; The determination of the fuel type of the fuel to be tested includes the following steps: The initial boiling point of the fuel to be tested is obtained during the micro-boiling range determination process; Retrieve and traverse the preset node database to determine whether there is a preset initial boiling point in the preset node database that matches the measured initial boiling point; If a preset initial boiling point exists that matches the measured initial boiling point, then the preset fuel type corresponding to the preset initial boiling point that matches the measured initial boiling point is taken as the fuel type of the fuel to be tested.

[0030] Specifically, the initial boiling point is a direct characterization of the volatility of light components in a fuel and is strongly correlated with the fuel type. Therefore, this embodiment uses the initial boiling point to determine the fuel type of the fuel to be tested, in order to improve detection accuracy and shorten response time (compared to manual input). The preset initial boiling point is a range of initial boiling point fluctuations (not a fixed value) for fuel types determined by statistical analysis of historical data, such as 140–160℃ for jet fuel. In the micro-boiling range test, when the first drop of condensate falls into the recovery unit (according to GB / T6536), the temperature sensor records and measures the initial boiling point in real time. Initial boiling point matching judgment: the preset node database is retrieved (structure example shown in Table 3), and the preset initial boiling point ranges of all preset fuel types are traversed. Matching rule: if the measured initial boiling point falls within the preset range of a certain type (e.g., 152℃ ∈ [140,160]), it is determined to be jet fuel.

[0031] In a preferred embodiment, the preset node database further includes standard gradient template vectors corresponding to each group of preset fuel types; After determining whether a preset initial boiling point matching the measured initial boiling point exists in the preset node database, the method further includes the following steps: If there is no preset initial boiling point that matches the measured initial boiling point, the steam temperature detection values ​​of at least three equally spaced nodes of the fuel to be tested are extracted in real time within the range from the initial boiling point to 5% recovery volume percentage to form a characteristic temperature sequence. Calculate the first-order difference gradient vector of the characteristic temperature sequence; Traverse the standard gradient template vectors corresponding to each group of the preset fuel types in the preset node database, and calculate the similarity between the first-order difference gradient vector and each of the standard gradient template vectors respectively; The preset fuel type corresponding to the standard gradient template vector that minimizes the similarity is selected as the fuel type of the fuel to be tested.

[0032] Specifically, the standard gradient template vector is obtained by pre-storing historical data, and the preset node database is shown in Table 3: Table 3 ; When the initial boiling point cannot be matched, the fuel type is identified through the boiling range gradient feature. The steps include: constructing a characteristic temperature sequence: selecting three nodes at equal intervals (e.g., 1.67%, 3.33%, 5%) from the initial boiling point to 5%, and recording the steam temperature values ​​in real time to form a characteristic temperature sequence S=[Ta,Tb,Tc]. For example, the characteristic temperature sequence of the jet fuel sample is [154.0℃,160.1℃,167.7℃]. The first-order difference gradient vector is calculated as follows: =[Tb−Ta,Tc−Tb], for example =[160.1−154.0,167.7−160.1]=[6.1,7.6], using the formula The similarity between the first-order difference gradient vector and each of the standard gradient template vectors is obtained, where G1 represents Tb−Ta and G2 represents Tc−Tb. Let represent the standard gradient template vector. For example, the jet fuel template [5.8, 7.2] is used. The similarity is calculated using the formula above. =0.5, which is the minimum, so the match is successful, and the type of fuel to be tested is jet fuel.

[0033] This implementation solves the problem of identifying unknown fuels, such as biodiesel (whose initial boiling point is 180°C, exceeding the preset range), by accurately classifying them through gradient matching; gradient features suppress single-point temperature errors and have strong anti-interference capabilities: the principle on which it is based is that the distillation rates of light fractions of different fuels are significantly different (5% of jet fuel distills within 5 minutes, while diesel requires 8 minutes), and gradient vectors quantify this kinetic characteristic.

[0034] Furthermore, the initial boiling point range may overlap between certain fuel types. In this case, to improve the accuracy of fuel type determination, the initial boiling point determination and gradient identification can be combined, and the two dimensions can be compared separately to determine a more accurate fuel type.

[0035] In a preferred embodiment, after obtaining the corresponding first corrected steam temperature, the method further includes the following step: Real-time acquisition of ambient atmospheric pressure; The method of predicting the closed-cup flash point by linear weighting of each of the first corrected steam temperatures includes the following steps: If the deviation between the ambient atmospheric pressure and the standard atmospheric pressure is less than or equal to the second preset threshold, the closed-cup flash point is predicted by linear weighted calculation of each of the first corrected vapor temperatures. After obtaining the real-time ambient atmospheric pressure, the following steps are also included: If the deviation between the ambient atmospheric pressure and the standard atmospheric pressure is greater than the second preset threshold, the first corrected steam temperature is corrected based on the deviation between the ambient atmospheric pressure and the standard atmospheric pressure to obtain the second corrected steam temperature. The closed-cup flash point is predicted by linear weighting of each of the second corrected steam temperatures.

[0036] Specifically, the current atmospheric pressure (unit: kPa) is read using a digital barometer (such as the BMP280 sensor). Pressure deviation judgment: The second preset threshold can be selected as 1 kPa (approximately corresponding to a change of 100 meters in altitude); if the absolute value of the difference between the ambient atmospheric pressure and the standard atmospheric pressure (101.3) is ≤1, the flash point is directly calculated using the first correction temperature; if the absolute value of the difference between the ambient atmospheric pressure and the standard atmospheric pressure (101.3) is >1, pressure compensation is activated using the formula... The second corrected steam temperature is obtained, where β represents the pressure-temperature compensation coefficient, determined through calibration experiments (typical value for jet fuel: 0.12℃ / kPa). The second corrected steam temperature is then used to replace the first corrected steam temperature in the weighting formula to obtain the predicted closed-cup flash point.

[0037] This implementation reduces errors in high-altitude scenarios and compensates for the decrease in boiling point caused by the decrease in air pressure; it is compatible with applications in a wider range of regions: covering altitudes of 0–4500 meters (air pressure 101.3–58 kPa). The principle it is based on is that boiling point and air pressure have a linear relationship. For every 1 kPa decrease, the boiling point decreases by about 0.3–0.5℃, and the temperature of light distillates needs to be corrected accordingly.

[0038] In a preferred embodiment, after acquiring the preceding distillation range temperature set in real time, the method further includes the following steps: The sliding standard deviation of the rate of change of steam temperature for adjacent preset recovery volume percentage nodes in the preceding distillation temperature group is calculated in real time. Specifically, the steam temperature change rate is the rate of temperature change between adjacent nodes (°C / %), and the sliding standard deviation is calculated in real time using a moving window of length 3 (example window [Δr1, Δr2, Δr3]), which is the standard deviation of the rate of change within the window, reflecting the data volatility.

[0039] The step of using the heat capacity perturbation factor to correct the initial steam temperature in each of the preceding distillation range temperature groups in real time to obtain the corresponding first corrected steam temperature includes the following steps: If the rate of change of steam temperature at all the preset recovery volume percentage nodes does not deviate from the current moving average by more than 3 times the moving standard deviation, then the heat capacity disturbance factor is used to correct the initial steam temperature in the preceding distillation temperature group in real time to obtain the corresponding first corrected steam temperature.

[0040] For example, for each new node (e.g., 4%), update the rate of change sequence (delete the oldest value and add the current value). Calculate the current moving mean μ and moving standard deviation σ. For example, if [Δr1=5.2, Δr2=5.5, Δr3=5.3], then the current moving mean μ=5.33 and the moving standard deviation σ=0.13 are calculated. Check whether the rate of change of all nodes satisfies |Δr i If -μ|≤3σ, then there are no outliers exceeding 3 times the standard deviation. In this case, the heat capacity perturbation factor HDF is called to calculate the corrected temperature in real time by step S2.

[0041] This implementation uses the 3σ criterion to eliminate transient disturbances such as bubble bursting (no correction is triggered when σ exceeds the limit), ensuring that compensation is only made for systematic heat capacity deviations (such as heater power drift).

[0042] In a preferred embodiment, after calculating the sliding standard deviation of the steam temperature change rate of adjacent preset recovery volume percentage nodes in the preceding distillation range temperature group in real time, the method further includes the following steps: If the steam temperature change rate at the preset recovery volume percentage node deviates from the current moving average by more than 3... Then, nodes with a steam temperature change rate that deviates from the current sliding mean by more than 3 times the sliding standard deviation of the preset recovery volume percentage are marked as abnormal nodes; The initial steam temperature corresponding to the abnormal node is corrected by using the linear interpolation result of the initial steam temperature of the two preset recovery volume percentage nodes adjacent to the abnormal node.

[0043] Specifically, the correction of abnormal nodes relies on the time-series data from the steam temperature detection unit, and linear interpolation utilizes the temperature values ​​of previous valid nodes (e.g., 1%, 2%). An abnormal node represents a steam temperature change rate that satisfies |Δr|. i Nodes where -μ|>3σ (e.g., 3% recovery volume) indicate a sensor interruption or localized boiling anomaly. Anomaly marking and handling: Identify anomalous nodes (e.g., when V=3%, Δr3=8.2, μ=5.3, σ=0.2, |8.2-5.3|=2.9>0.6). Linear interpolation correction: Take the first two valid nodes (V... a =1%, T a =152℃; V b =2%, T b =158℃), calculate the linear interpolation results. Replace the original measurement value of the abnormal node (e.g., the original value of the 3% node is 170℃, which is corrected to 164℃).

[0044] In a preferred embodiment, after calculating the sliding standard deviation of the steam temperature change rate of adjacent preset recovery volume percentage nodes in the preceding distillation range temperature group in real time, the method further includes the following steps: If two or more of the preset recovery volume percentage nodes are marked as abnormal nodes, the liquid temperature detection values ​​at the time of the abnormal node corresponding to all the abnormal nodes are extracted in real time. Furthermore, after extracting the liquid temperature detection values ​​at the time of the anomaly corresponding to all the anomaly nodes in real time, the following steps are also included: Determine whether the temperature gradient of the liquid exceeds the temperature change threshold; Based on the liquid temperature detection value at the time of the abnormal node and the temperature offset of the heat capacity disturbance factor, the compensation steam temperature corresponding to each abnormal node is obtained. Furthermore, when the liquid temperature change gradient does not exceed the temperature change threshold, the compensated steam temperature corresponding to each abnormal node is obtained based on the liquid temperature detection value at the time of the abnormal node and the temperature offset of the heat capacity perturbation factor. This step takes into account that the liquid temperature detection unit is located inside the distillation flask. If the steam temperature is abnormal (such as bubble rupture), the liquid temperature may also be distorted synchronously. Compensating the steam temperature with an unreliable liquid temperature may amplify the error. Therefore, determining the authenticity of the liquid temperature before compensation can improve accuracy.

[0045] The initial steam temperature is corrected using the compensation steam temperature corresponding to each of the aforementioned abnormal nodes.

[0046] Specifically, the liquid temperature at the time of the abnormal node represents the measured value of the liquid temperature detection unit at the timestamp corresponding to the abnormal node. The compensated steam temperature represents the steam phase temperature calculated based on the liquid temperature and the heat capacity disturbance factor. The compensation mechanism is triggered when two adjacent nodes (e.g., 3% and 4%) are both marked as abnormal. Specifically: Liquid temperature extraction: The liquid temperature at the time of the abnormal node (e.g., T) is read from the device database. {liq,3%} =145℃, T {liq,4%} =150℃). Compensated steam temperature calculation: Obtain the temperature offset of the heat capacity disturbance factor. Where 160 is a calibration constant, a conversion coefficient determined through heat transfer experiments; for example, HDF = 0.15, temperature offset is 24℃, and compensated steam temperature is... =145 + 0.8 × 24 = 164.2℃, where, The heat transfer coefficient, calibrated through boiling point experiments, can be selected as 0.8. Replace the initial steam temperature at abnormal nodes with the compensated steam temperature (e.g., the original value of 170℃ at the 3% node is corrected to 164.2℃). Therefore, this prediction method demonstrates good equipment compatibility and ease of implementation. Primarily based on software algorithm innovation, it can be easily integrated into the control system of existing micro-range distillation devices. Only the addition of corresponding data acquisition modules (such as pressure sensors) and calculation software is required, without significant hardware modifications, demonstrating excellent compatibility and widespread applicability. Furthermore, it provides additional valuable insights: real-time flash point prediction during distillation range testing helps operators understand the sample's safety characteristics (volatility) earlier and adjust subsequent testing strategies or safety measures as needed. Simultaneously, the synchronously acquired high-precision distillation range and flash point data provide a more consistent and reliable data foundation for in-depth research into the intrinsic relationship between oil composition, volatility, and safety. In summary, this application seamlessly embeds closed-cup flash point prediction into the micro-range distillation process. Through precise selection of relevant nodes, real-time compensation for thermal disturbances, and optimized weighted calculations, it achieves high-precision flash point prediction values ​​even before the distillation range test is completed. This solves the problems of test separation, low efficiency, and poor data synergy in traditional methods, providing a new, efficient, accurate, and economical approach for petroleum product analysis.

[0047] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. The above descriptions are only preferred embodiments of this application. It should be noted that due to the limitations of written expression, while there are objectively infinite specific structures, those skilled in the art can make several improvements, modifications, or changes without departing from the principles of this application, and can also combine the above technical features in an appropriate manner. These improvements, modifications, changes, or combinations, or the direct application of the inventive concept and technical solution to other situations without modification, should all be considered within the scope of protection of this application.

Claims

1. A method for real-time prediction of closed-cup flash point based on a micro-distillation range measuring device, characterized in that, Includes the following steps: The preceding distillation range temperature group is acquired in real time. The preceding distillation range temperature group includes the initial vapor temperature corresponding to multiple preset recovery volume percentage nodes. The preset recovery volume percentage nodes are located in the initial rising segment of the boiling range curve of the micro-distillation range measuring device, and the linear correlation coefficient with the closed-cup flash point is greater than or equal to a first preset threshold. Real-time acquisition of steam pressure and liquid temperature is used to obtain the rate of change of steam pressure and the gradient of change of liquid temperature, thereby calculating the thermal capacity disturbance factor of the device. ； Wherein, HDF represents the heat capacity perturbation factor, k1 and k2 are weighting coefficients determined through calibration experiments, ΔP / Δt represents the rate of change of steam pressure, and ΔT / Δt represents the gradient of change of liquid temperature. The initial steam temperature in each of the preceding distillation range temperature groups is corrected in real time using the heat capacity perturbation factor to obtain the corresponding first corrected steam temperature: ； Among them, T 修正 T represents the first corrected steam temperature. 初始 Indicates the initial steam temperature. α This represents the temperature correction factor, obtained through thermodynamic calibration. Based on the pre-established weighting relationship of the linear correlation between the test steam temperature and the closed-cup flash point of the preset recovery volume percentage node, the closed-cup flash point is predicted by linear weighting of each of the first corrected steam temperatures.

2. The method for real-time prediction of closed-cup flash point based on a micro-distillation range measuring device according to claim 1, characterized in that: The real-time acquisition of the preceding distillation range temperature group includes the following steps: Determine the fuel type of the fuel to be tested; Based on the fuel type, a corresponding preset node set is obtained, and the preset node set includes multiple preset recycling volume percentage nodes; During the micro-distillation range determination process, when the recovery volume percentage reaches each preset recovery volume percentage node, the initial steam temperature measured by the steam temperature detection unit is recorded in real time to form a preceding distillation range temperature group.

3. The method for real-time prediction of closed-cup flash point based on a micro-distillation range measuring device according to claim 2, characterized in that: The step of obtaining the corresponding preset node set based on the fuel type includes the following steps: The preset node database is retrieved and traversed to obtain a preset node set corresponding to the fuel type of the fuel to be tested; the preset node database includes multiple preset fuel types and preset node sets corresponding to each fuel type; The method also includes constructing the preset node database; The construction of the preset node database includes the following steps: Multiple groups of fuel samples were selected, and the distillation range test of the standard distillation range test method was performed on each group of fuel samples. The test vapor temperature of all test nodes between the initial boiling point and 10% recovery volume percentage was recorded. Simultaneously perform closed-cup flash point testing using the standard closed-cup flash point test method on each group of fuel samples to obtain the test closed-cup flash point; For each group of fuel samples, calculate the linear correlation coefficient between the test vapor temperature and the test closed-cup flash point for each test node of the fuel sample in that group; The test nodes whose linear correlation coefficient is greater than or equal to the first preset threshold are retained as preset recovery volume percentage nodes corresponding to the group of fuel samples, thereby obtaining the preset node set corresponding to the group of fuel samples; Wherein, the first preset threshold is greater than or equal to 0.

99.

4. The method for real-time prediction of closed-cup flash point based on a micro-distillation range measuring device according to claim 2, characterized in that: The fuel type includes jet fuel or diesel; The step of obtaining the corresponding preset node set based on the fuel type includes the following steps: If the fuel type is jet fuel or diesel, the corresponding preset node set includes four preset recovery volume percentage nodes of 1%, 2%, 3% and 4%.

5. The method for real-time prediction of closed-cup flash point based on a micro-distillation range measuring device according to claim 3, characterized in that: The preset node database also includes the preset initial boiling point corresponding to each group of preset fuel types; The determination of the fuel type of the fuel to be tested includes the following steps: The initial boiling point of the fuel to be tested is obtained during the micro-boiling range determination process; Retrieve and traverse the preset node database to determine whether there is a preset initial boiling point in the preset node database that matches the measured initial boiling point; If a preset initial boiling point exists that matches the measured initial boiling point, then the preset fuel type corresponding to the preset initial boiling point that matches the measured initial boiling point is taken as the fuel type of the fuel to be tested.

6. The method for real-time prediction of closed-cup flash point based on a micro-distillation range measuring device according to claim 5, characterized in that: The preset node database also includes standard gradient template vectors corresponding to each group of preset fuel types; After determining whether a preset initial boiling point matching the measured initial boiling point exists in the preset node database, the method further includes the following steps: If there is no preset initial boiling point that matches the measured initial boiling point, the steam temperature detection values ​​of at least three equally spaced nodes of the fuel to be tested are extracted in real time within the range from the initial boiling point to 5% recovery volume percentage to form a characteristic temperature sequence. Calculate the first-order difference gradient vector of the characteristic temperature sequence; Traverse the standard gradient template vectors corresponding to each group of the preset fuel types in the preset node database, and calculate the similarity between the first-order difference gradient vector and each of the standard gradient template vectors respectively; The preset fuel type corresponding to the standard gradient template vector that minimizes the similarity is selected as the fuel type of the fuel to be tested.

7. The method for real-time prediction of closed-cup flash point based on a micro-distillation range measuring device according to claim 1, characterized in that: After obtaining the corresponding first corrected steam temperature, the method further includes the following steps: Real-time acquisition of ambient atmospheric pressure; The method of predicting the closed-cup flash point by linear weighting of each of the first corrected vapor temperatures includes the following steps: If the deviation between the ambient atmospheric pressure and the standard atmospheric pressure is less than or equal to the second preset threshold, the closed-cup flash point is predicted by linear weighted calculation of each of the first corrected vapor temperatures. After obtaining the real-time ambient atmospheric pressure, the following steps are also included: If the deviation between the ambient atmospheric pressure and the standard atmospheric pressure is greater than the second preset threshold, the first corrected steam temperature is corrected based on the deviation between the ambient atmospheric pressure and the standard atmospheric pressure to obtain the second corrected steam temperature. The closed-cup flash point is predicted by linear weighting of the second corrected steam temperatures.

8. The method for real-time prediction of closed-cup flash point based on a micro-distillation range measuring device according to claim 1, characterized in that: After acquiring the preceding distillation range temperature set in real time, the following steps are also included: The sliding standard deviation of the rate of change of steam temperature for adjacent preset recovery volume percentage nodes in the preceding distillation temperature group is calculated in real time. The step of using the heat capacity perturbation factor to correct the initial steam temperature in each of the preceding distillation range temperature groups in real time to obtain the corresponding first corrected steam temperature includes the following steps: If the rate of change of steam temperature at all the preset recovery volume percentage nodes does not deviate from the current moving average by more than 3 times the moving standard deviation, then the heat capacity disturbance factor is used to correct the initial steam temperature in the preceding distillation temperature group in real time to obtain the corresponding first corrected steam temperature.

9. The method for real-time prediction of closed-cup flash point based on a micro-distillation range measuring device according to claim 8, characterized in that: After calculating the sliding standard deviation of the steam temperature change rate of adjacent preset recovery volume percentage nodes in the preceding distillation range temperature group in real time, the method further includes the following steps: If the steam temperature change rate of the preset recovery volume percentage node deviates from the current moving average by more than 3 times the moving standard deviation, then the preset recovery volume percentage node whose steam temperature change rate deviates from the current moving average by more than 3 times the moving standard deviation is marked as an abnormal node. The initial steam temperature corresponding to the abnormal node is corrected by using the linear interpolation result of the initial steam temperature of the two preset recovery volume percentage nodes adjacent to the abnormal node.

10. The method for real-time prediction of closed-cup flash point based on a micro-distillation range measuring device according to claim 9, characterized in that: After calculating the sliding standard deviation of the steam temperature change rate of adjacent preset recovery volume percentage nodes in the preceding distillation range temperature group in real time, the method further includes the following steps: If two or more of the preset recovery volume percentage nodes are marked as abnormal nodes, the liquid temperature detection values ​​at the time of the abnormal node corresponding to all the abnormal nodes are extracted in real time. Based on the liquid temperature detection value at the time of the abnormal node and the temperature offset of the heat capacity disturbance factor, the compensation steam temperature corresponding to each abnormal node is obtained. The initial steam temperature corresponding to each abnormal node is corrected using the compensation steam temperature corresponding to each abnormal node.

Images