A method and system for temperature control of a gas boiler

Abstract

The application relates to the technical field of gas boiler control, in particular to a gas boiler temperature control method and system, which comprises the following steps: collecting raw material hyperspectral data of a heating and stirring furnace in a lubricating oil preparation process at each moment; obtaining a raw material heating uniformity change coefficient at each moment according to the differences in band reflectivity and spatial distribution characteristics of pixels in the raw material hyperspectral data at each moment; obtaining a heating and stirring furnace temperature adaptation degree by combining the raw material heating uniformity and the raw material heating uniformity change coefficient; and controlling the temperature of the heating and stirring furnace according to the numerical distribution of the heating and stirring furnace temperature adaptation degrees at all moments before the current moment. The application can realize intelligent control of the gas boiler temperature in the lubricating oil preparation process and improve the temperature control precision.

Description

Technical Field

[0001] This application relates to the field of gas boiler control technology, specifically to a gas boiler temperature control method and system. Background Technology

[0002] The preparation of lubricating oil mainly involves adding auxiliary materials such as rust inhibitors and friction modifiers to base oil, and then heating and stirring it using a gas-fired boiler until it is uniformly mixed. However, when using a traditional gas-fired boiler for heating, because the heating device of the heating and stirring furnace is located at the bottom of the furnace, heat is transferred from bottom to top. The raw material temperature in some areas is lower, resulting in weaker fluidity. This uneven heat transfer of the lubricating oil raw materials causes heat loss, leading to the problem that the upper layer of raw materials receives less heat and is heated unevenly. Consequently, the temperature control accuracy of the gas-fired boiler during the raw material stirring process is poor, resulting in a lower quality lubricating oil. Summary of the Invention

[0003] To address the aforementioned technical problems, the purpose of this application is to provide a method and system for temperature control of a gas-fired boiler, the specific technical solution of which is as follows:

[0004] In a first aspect, embodiments of this application provide a method for controlling the temperature of a gas-fired boiler, the method comprising the following steps:

[0005] S1, Collect the raw material hyperspectral data of the heating process of lubricating oil preparation in the heating and stirring furnace at each time point; the raw material hyperspectral data includes: the band reflectance of each pixel position to all bands;

[0006] S2, based on the differences in band reflectance and spatial distribution characteristics of pixels in the historical hyperspectral data of the raw materials at each time point, the uniformity of heating of the raw materials at each time point is obtained, specifically including:

[0007] S21. Based on the differences in band reflectance of pixels in the hyperspectral data of raw materials at each time point and the spatial distribution characteristics of pixels, the density of raw material distribution at each time point is obtained.

[0008] S22. Based on the differences in band reflectance and spatial distribution characteristics of all pixels in the hyperspectral data of the raw materials at adjacent time points, the stirring fluidity coefficient of the raw materials at each time point is obtained.

[0009] S23, combining the raw material stirring fluidity coefficient and raw material distribution density at each moment, to obtain the raw material heating uniformity;

[0010] S24. Based on the difference in the degree of disorder of the distribution of band reflectance of all pixels in the hyperspectral data of raw materials at adjacent time points, the raw material heating and stirring speed at each time point is obtained; based on the degree of dispersion of the trend of the change of the raw material heating and stirring speed at each time point, the raw material temperature uniformity index at each time point is obtained; based on the numerical distribution of the raw material temperature uniformity index at all time points, the raw material heating uniformity variation coefficient at each time point is obtained.

[0011] S3, combining the uniformity of raw material heating and the uniformity of raw material heating variation coefficient, obtains the temperature adaptability of the heating and stirring furnace. Based on the numerical distribution of the temperature adaptability of the heating and stirring furnace at all times before the current time, the temperature of the heating and stirring furnace is controlled.

[0012] Furthermore, the method for obtaining the raw material distribution density at each time point includes:

[0013] The sequence of band reflectance of each pixel's corresponding position for all bands is denoted as the spectral sequence of each pixel; the distance between the spectral sequences of any two pixels is used as the metric distance, and density clustering is performed on all pixels to obtain each raw material cluster;

[0014] For each raw material cluster, the Euclidean distance between any two pixels is used as the metric distance. Density clustering is performed on all pixels contained within the raw material cluster to obtain each sub-cluster.

[0015] For each raw material cluster, the pixel with the smallest x-coordinate in each sub-cluster of the raw material cluster is obtained and denoted as the representative pixel of each sub-cluster; for any two sub-clusters, the Euclidean distance between the representative pixels of the two sub-clusters is calculated and denoted as the first distance; the mean of the first distances of all sub-clusters contained in the raw material cluster is obtained; the ratio of the total number of pixels in the raw material cluster to the mean is taken as the raw material aggregation degree of each raw material cluster.

[0016] Obtain the sum of the mean values ​​of all raw material clusters, and record it as the total raw material density value; calculate the cumulative sum of the raw material aggregation degree of all raw material clusters, and use the ratio of the cumulative sum to the total raw material density value as the raw material distribution density at each time point.

[0017] Furthermore, the method for obtaining the material stirring fluidity coefficient at each time point includes:

[0018] The maximum value of the band reflectance of each pixel's corresponding position for all bands is selected as the representative reflectance value of each pixel; the Euclidean distance between any two pixels within each raw material cluster is obtained; the average value of all the Euclidean distances within each raw material cluster is obtained.

[0019] For the raw material hyperspectral data at each time point, calculate the information entropy of the representative reflectance values ​​of all pixels in the raw material hyperspectral data; obtain the sum of the average values ​​of all raw material clusters in the raw material hyperspectral data, and record it as the discrete total value of the raw material at each time point;

[0020] Based on the differences in the numerical distribution of representative reflectance values ​​of all pixels in the hyperspectral data of the raw material at adjacent time points, as well as the differences in the total discrete value of the raw material, the stirring fluidity coefficient of the raw material at each time point is obtained.

[0021] Furthermore, the method for obtaining the raw material stirring fluidity coefficient at each time point includes:

[0022] For each time other than the initial time, calculate the difference in information entropy between each time and the previous time, and denot it as information difference; obtain the difference in the total discrete value of raw materials between each time and the previous time, and denot it as raw material distribution difference; and fused the raw material distribution difference and information difference between each time and the previous time, and denot it as the raw material mixing fluidity coefficient at each time.

[0023] Furthermore, the step of combining the raw material stirring fluidity coefficient and the raw material distribution density at each time point to obtain the raw material heating uniformity includes: using the ratio of the raw material stirring fluidity coefficient to the raw material distribution density at each time point as the raw material heating uniformity at each time point.

[0024] Furthermore, the method for obtaining the raw material heating and stirring speed at each time point includes:

[0025] Calculate the information entropy of the representative reflectance values ​​of all pixels in the hyperspectral data of the raw material at each time point, and record it as the discrete value of the raw material at each time point; record the ratio of the difference between the discrete value of the raw material at each time point and the previous time point to the preset sampling interval T as the heating and stirring rate of the raw material at each time point.

[0026] Furthermore, the method for obtaining the raw material temperature equilibrium index at each time point specifically includes:

[0027] The sequence of raw material heating and stirring speeds at a predetermined number of time points prior to each time point is denoted as the trend sequence at each time point.

[0028] Calculate the sum of the average values ​​of all raw material clusters at each time point, and record it as the raw material temperature discrete value at each time point; calculate the difference between the raw material temperature discrete value at each time point and the previous time point, and record it as the first difference; calculate the variance of the trend sequence at each time point, and use the ratio of the first difference to the variance as the raw material temperature equilibrium index at each time point.

[0029] Furthermore, the method for obtaining the uniformity variation coefficient of raw material heating at each time point includes:

[0030] Obtain the trend strength of the trend sequence at each time point, denoted as the raw material mixing trend coefficient at each time point; perform threshold segmentation on the raw material mixing trend coefficients at all times points to obtain the mixing judgment threshold; if the raw material mixing trend coefficient at each time point is greater than or equal to the mixing judgment threshold, then the raw material heating uniformity change coefficient at each time point is the raw material temperature equilibrium index at each time point; otherwise, the raw material heating uniformity change coefficient at each time point is the reciprocal of the raw material temperature equilibrium index at each time point.

[0031] Furthermore, the temperature adaptability of the heating and stirring furnace is obtained by combining the uniformity of raw material heating and the coefficient of variation of raw material heating uniformity. Based on the numerical distribution of the temperature adaptability of the heating and stirring furnace at all times prior to the current moment, the temperature of the heating and stirring furnace is controlled, including:

[0032] The product of the uniformity of raw material heating and the coefficient of variation of raw material heating uniformity at each time moment is taken as the temperature adaptability of the heating and stirring furnace at each time moment.

[0033] A prediction algorithm is used to obtain the heating and stirring furnace temperature fit at the current moment by considering the heating and stirring furnace temperature fit at all moments before the current moment; threshold segmentation is performed on the heating and stirring furnace temperature fit at all moments to obtain the optimal segmentation threshold.

[0034] If the temperature fit of the heating and stirring furnace at the current moment is less than or equal to the optimal segmentation threshold, then adjust the temperature of the heating and stirring furnace.

[0035] Secondly, embodiments of this application also provide a gas-fired boiler temperature control system for lubricating oil preparation, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described gas-fired boiler temperature control methods.

[0036] This application has at least the following beneficial effects:

[0037] This application analyzes the hyperspectral data distribution characteristics and band reflectance differences at various moments during the heating, stirring, and mixing process of raw materials in lubricating oil preparation to obtain the raw material distribution density at each moment, accurately reflecting the distribution characteristics of different raw materials. Furthermore, it analyzes the changes in data distribution characteristics and band reflectance differences between adjacent moments to obtain the raw material stirring fluidity coefficient, reflecting the gradual change in raw material distribution during the stirring process. Combining the raw material stirring fluidity coefficient and raw material distribution density, the raw material heating uniformity is obtained, reflecting the overall uniformity of heating at a single moment. Further, based on the differences in heating uniformity at historical moments combined with changes in stirring speed, a raw material heating uniformity variation coefficient is obtained, more accurately reflecting the uniformity of heating changes during the heating and stirring process. Furthermore, a heating and stirring furnace temperature adaptability is constructed to characterize the suitability of the current heating temperature in the heating and stirring furnace, using the heating temperature suitability to determine whether the heating temperature needs adjustment. This application solves the problem of poor temperature control accuracy of gas-fired boilers caused by uneven heating of raw materials during stirring and heating in the lubricating oil preparation process. Attached Figure Description

[0038] To more clearly illustrate the technical solutions and advantages 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, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 A flowchart illustrating the steps of a gas-fired boiler temperature control method according to one embodiment of this application;

[0040] Figure 2 This is a schematic diagram illustrating a scenario for acquiring hyperspectral data of raw materials according to an embodiment of this application.

[0041] Figure 3 This is a block diagram illustrating the acquisition of the uniformity variation coefficient of raw material heating at various times, as provided in one embodiment of this application. Detailed Implementation

[0042] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. In the absence of conflict, the embodiments of this application and the technical features in the embodiments can be combined with each other. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0043] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0044] The following description, in conjunction with the accompanying drawings, details a specific scheme for a gas-fired boiler temperature control method and system provided in this application.

[0045] Please see Figure 1 The diagram illustrates a flowchart of a gas boiler temperature control method according to an embodiment of this application, which includes the following steps:

[0046] S1, collect the raw material hyperspectral data of the heating process of lubricating oil preparation in the heating and stirring furnace at various times.

[0047] One specific implementation scenario of this application is the temperature control of a gas-fired boiler during the heating and preparation of lubricating oil. In this scenario, uneven heat transfer within the heating and stirring furnace during the lubricating oil raw material heating process leads to lower heat distribution to the upper layer of raw material, resulting in uneven heating, poor material flowability, and ultimately, poor quality lubricating oil. Therefore, real-time monitoring of the temperature data within the heating and stirring furnace is necessary.

[0048] Currently, the method for monitoring the temperature of gas-fired boilers involves analyzing the spectral data obtained from spectral imaging to interpret the actual heating conditions of the raw materials during the heating and stirring process. However, in the process of lubricating oil preparation, the heating and stirring in the furnace are carried out gradually. Uneven heating leads to relatively poor flowability of the raw materials, resulting in a slow dispersion rate of the raw materials under stirring, which affects the real-time judgment of the heating status based on spectral data.

[0049] Therefore, in the embodiments of this application, such as Figure 2 As shown, a hyperspectral camera A is positioned directly above the infrared window B of the heating and stirring furnace C to capture images vertically. Hyperspectral data of the lubricating oil inside the furnace is acquired every 2 seconds (T=2s), and noise is removed using mean filtering. Mean filtering is a known technique, and its process will not be described in detail here. Taking a single moment as an example, the hyperspectral data of the raw material includes the band reflectance of each pixel's corresponding position across all bands.

[0050] The heating temperature of the heating and stirring furnace can be adjusted based on the numerical distribution of the representative reflectance values ​​of pixels at different times and locations.

[0051] S2, based on the differences in band reflectance and spatial distribution characteristics of pixels in the historical hyperspectral data of raw materials at each time point, the uniformity of heating of raw materials at each time point is obtained as the uniformity of heating of raw materials at each time point.

[0052] During the heating process of lubricating oil raw materials, the fluidity of the raw materials increases. However, when the raw materials are heated unevenly, the temperature of the raw materials in some areas may be lower, resulting in weaker fluidity. During the stirring process, the raw materials in some areas are less dispersed under the same stirring degree. In the hyperspectral image, the positional change of the raw materials at this location is not obvious. The raw materials that are heated unevenly will gather in the uneven area, while a small part is more dispersed.

[0053] S21. Based on the differences in band reflectance of pixels in the hyperspectral data of raw materials at each time point and the spatial distribution characteristics of pixels, the density of raw material distribution at each time point is obtained.

[0054] Based on the above analysis, the unevenly heated raw materials in the heating and stirring furnace will converge in the uneven areas, while a small portion will be more dispersed, exhibiting different distribution characteristics. Furthermore, the uneven raw materials show varying degrees of spectral reflectance at different locations. Therefore, the density of the raw material distribution can be determined based on the differences in spectral reflectance across all wavelengths and the spatial distribution characteristics of the pixels.

[0055] First, the sequence of band reflectances of each pixel's corresponding position for all bands is denoted as the spectral sequence of each pixel; the distance between the spectral sequences of any two pixels is used as the metric distance, and density clustering is performed on all pixels to obtain each raw material cluster;

[0056] For each raw material cluster, the Euclidean distance between any two pixels is used as the metric distance. Density clustering is performed on all pixels contained within the raw material cluster to obtain each sub-cluster.

[0057] It should be noted that density clustering is an algorithm that clusters data points based on density structure without requiring a pre-specified number of clusters. Implementers can choose density clustering algorithms such as DPC clustering, DBSCAN clustering, and OPTICS clustering according to their actual situation, and this application does not impose any restrictions. In this embodiment, the distance between the spectral sequences of any two pixels is used as the metric distance. All pixels are clustered using the DPC clustering algorithm, with the truncation distance set to 3. The metric distance is the distance between the spectral sequences of the pixels. Here, the distance between two sequences represents the difference between the two sequences. Implementers can choose dtw distance, Euclidean distance, Manhattan distance, etc., according to their actual situation, and this application does not impose any restrictions.

[0058] In this embodiment, the dtw distance is selected as the distance metric between spectral sequences of pixels; furthermore, the DBSCAN clustering algorithm is used to perform density clustering on all pixels contained within the raw material cluster. The DPC clustering algorithm, the DBSCAN clustering algorithm, and the calculation of the dtw distance are all well-known techniques, and their processes will not be described in detail here.

[0059] Each cell in a raw material cluster represents a single raw material. Taking a raw material cluster as an example, the Euclidean distance between all cells in that cluster is calculated. Using the DBSCAN clustering algorithm, with a minimum density of 5 and a neighborhood radius of 3, the output consists of multiple sub-clusters.

[0060] For each raw material cluster, the pixel with the smallest x-coordinate in each sub-cluster of the raw material cluster is obtained and denoted as the representative pixel of each sub-cluster; for any two sub-clusters, the Euclidean distance between the representative pixels of the two sub-clusters is calculated and denoted as the first distance; the mean of the first distance of all sub-clusters contained in the raw material cluster is obtained; the ratio of the total number of pixels in the raw material cluster to the mean is taken as the raw material aggregation degree of each raw material cluster.

[0061] Obtain the sum of the mean values ​​of all raw material clusters, and record it as the total raw material density value; calculate the cumulative sum of the raw material aggregation degree of all raw material clusters, and use the ratio of the cumulative sum to the total raw material density value as the raw material distribution density at each time point.

[0062] The lower the density of raw material distribution at a single moment, the greater the uniformity of heating of the raw materials, indicating that the distribution of each raw material is relatively discrete, the overall heating of the raw materials is relatively uniform, and the mixing effect is better.

[0063] S22. Based on the differences in band reflectance and spatial distribution characteristics of all pixels in the hyperspectral data of the raw materials at adjacent time points, the stirring fluidity coefficient of the raw materials at each time point is obtained.

[0064] Since the reflectivity of different positions in the heating and stirring furnace varies greatly for different wavelengths during the stirring process, in order to more clearly reflect the temperature distribution at different positions, the positional distribution of the unevenly heated lubricating oil raw material in the heating and stirring furnace at adjacent times is combined to construct the raw material stirring fluidity coefficient at each time.

[0065] The maximum value of the band reflectance of each pixel's corresponding position for all bands is selected as the representative reflectance value of each pixel; the Euclidean distance between any two pixels within each raw material cluster is obtained; the average value of all the Euclidean distances within each raw material cluster is obtained.

[0066] For the raw material hyperspectral data at each time point, calculate the information entropy of the representative reflectance values ​​of all pixels in the raw material hyperspectral data; obtain the sum of the average values ​​of all raw material clusters in the raw material hyperspectral data, and record it as the discrete total value of the raw material at each time point;

[0067] For each time other than the initial time, calculate the difference in information entropy between each time and the previous time, and denot it as information difference; obtain the difference in the total discrete value of raw materials between each time and the previous time, and denot it as raw material distribution difference; and fused the raw material distribution difference and information difference between each time and the previous time, and denot it as the raw material mixing fluidity coefficient at each time.

[0068] It should be noted that fusion indicates a unidirectional relationship between the fused data and the fused result, where the larger the fused data, the larger the fused result. Fusion can specifically involve multiplication, addition, etc., determined by the actual application; this application does not impose any special restrictions. Difference represents the degree of difference between two values; in this embodiment, the absolute value of the difference is used to calculate the difference.

[0069] As an embodiment of this application, the material stirring fluidity coefficient at each moment is the product of the material distribution difference and information difference between each moment and the previous moment;

[0070] As another embodiment of this application, the material mixing fluidity coefficient at each time point is the sum of the material distribution differences and information differences between each time point and the previous time point.

[0071] The greater the material flowability coefficient at a single moment, the greater the uniformity of heating of the material. This indicates that the distribution of all material pixels is more chaotic at the current moment compared to the previous moment, and the faster the position changes due to stirring, the stronger the flowability and the higher the uniformity of heating.

[0072] S23, combining the raw material stirring fluidity coefficient and raw material distribution density at each time point, obtains the raw material heating uniformity.

[0073] The material mixing fluidity coefficient at a single moment can be determined by the material pixel disorder and the rate of change of material distribution at that single moment.

[0074] Specifically, the ratio of the raw material stirring fluidity coefficient to the raw material distribution density at each time point is used as the raw material heating uniformity at each time point.

[0075] The greater the uniformity of heating of the raw material at a single moment, the more dispersed the distribution of the lubricating oil raw material at that moment, and the stronger the fluidity of the raw material. The raw material is heated more evenly during the heating process, the better the heating effect of the heating and stirring furnace, the more suitable the temperature, and the less need there is for temperature control of the heating and stirring furnace.

[0076] S24. Based on the difference in the degree of disorder in the distribution of band reflectance of all pixels in the hyperspectral data of raw materials at adjacent time points, the heating and stirring speed of raw materials at each time point is obtained.

[0077] In the lubricating oil preparation process, the heating and stirring in the heating furnace are carried out gradually. In the initial heating stage, the temperature of the lubricating oil raw materials has not fully risen, and their fluidity is relatively poor. Therefore, the raw materials disperse slowly under stirring. Furthermore, the heating in the heating furnace is from bottom to top, so the raw materials in the upper layer of the furnace will be heated more slowly than those in the lower layer, requiring a certain amount of stirring time to achieve uniform heating. Therefore, controlling the temperature of the heating furnace solely based on the uniformity of raw material heating may lead to excessive temperature rise, thus affecting the quality of the prepared lubricating oil. Therefore, it is necessary to analyze the heating process during preparation.

[0078] Calculate the information entropy of the representative reflectance values ​​of all pixels in the hyperspectral data of the raw material at each time point, and record it as the discrete value of the raw material at each time point; record the ratio of the difference between the discrete value of the raw material at each time point and the previous time point to the preset sampling interval T as the heating and stirring rate of the raw material at each time point.

[0079] The sequence of raw material heating and stirring speeds for a preset number of M=10 time points prior to each time point is denoted as the trend sequence for each time point; the trend intensity of the trend sequence is obtained and denoted as the raw material mixing trend coefficient for each time point.

[0080] It should be noted that the trend strength of the sequence can be calculated using the trend strength formula in the STL sequence decomposition algorithm. This algorithm is a well-known technique, and its process will not be described in detail here. When there are fewer than 10 time points before the current time point, the data is filled using the mean filling method. Mean filling is a well-known technique, and its process will not be described in detail here.

[0081] A larger raw material mixing trend coefficient at a single moment indicates a stronger increasing trend in the raw material heating and stirring rate, suggesting it is more likely to be in the initial heating and stirring stage. Conversely, a smaller change in the raw material heating and stirring rate suggests it may be in the heat preservation stage after heating.

[0082] The raw material temperature equilibrium index for a single moment is determined based on the average change in distance between all pixels of each raw material at a single moment and the state of steady change.

[0083] Specifically, the sum of the average values ​​of all raw material clusters at each time point is calculated and denoted as the raw material temperature discrete value at each time point; the difference between the raw material temperature discrete value at each time point and the previous time point is calculated and denoted as the first difference; the variance of the trend sequence at each time point is calculated, and the ratio of the first difference to the variance is used as the raw material temperature equilibrium index at each time point.

[0084] The larger the raw material temperature uniformity index at a single moment, and the corresponding moment is the initial heating stage of raw material heating and stirring, the larger the corresponding raw material heating uniformity variation coefficient, indicating that the temperature change of the raw material during the heating process is relatively uniform up to the current moment, and the current heating temperature of the heating and stirring furnace is relatively suitable. Conversely, the smaller the raw material temperature uniformity index at a single moment, and the corresponding moment is the heat preservation stage of raw material heating and stirring, the larger the corresponding raw material heating uniformity variation coefficient, indicating that the temperature of the raw material heating and stirring is more uniform at this time.

[0085] Furthermore, based on the above analysis, the influence of the change in raw material temperature equilibrium index on the temperature uniformity of raw material heating and stirring is determined by the raw material mixing trend coefficient at a single moment and the raw material temperature equilibrium index, and the raw material heating uniformity variation coefficient is determined.

[0086] The degree of raw material mixing is divided into two situations: in the initial stage of mixing, the degree of raw material mixing increases; when mixing is basically completed, the degree of raw material mixing remains unchanged; further, the initial stage of mixing and the stage when mixing is completed are distinguished based on the changes in the uniformity of heating of the raw materials.

[0087] Specifically, the raw material mixing trend coefficient at all times is divided into thresholds to obtain the mixing judgment threshold. If the raw material mixing trend coefficient at each time is greater than or equal to the mixing judgment threshold, then the raw material heating uniformity change coefficient at each time is the raw material temperature equilibrium index at each time. Otherwise, the raw material heating uniformity change coefficient at each time is the reciprocal of the raw material temperature equilibrium index at each time.

[0088] In this embodiment, the raw material mixing trend coefficients at all times are used as input, and the Otsu threshold method is used to output the raw material mixing process judgment threshold. The Otsu threshold method is a well-known technique, and its process will not be described in detail here. Implementers can choose the threshold segmentation method according to actual conditions; this application does not impose any restrictions.

[0089] The larger the uniformity coefficient of raw material heating at a single moment, the more uniform the temperature rise of the raw material is during the heating process, and the more suitable the heating temperature of the heating furnace is.

[0090] A flowchart for obtaining the coefficient of variation of raw material heating uniformity at various times, as shown below. Figure 3 As shown

[0091] S3, combining the uniformity of raw material heating and the uniformity of raw material heating variation coefficient, obtains the temperature adaptability of the heating and stirring furnace. Based on the numerical distribution of the temperature adaptability of the heating and stirring furnace at all times before the current time, the temperature of the heating and stirring furnace is controlled.

[0092] The temperature adaptability of the heating and stirring furnace is constructed based on the uniformity of raw material heating at a single moment and the coefficient of uniform change of raw material heating.

[0093] The product of the uniformity of raw material heating and the coefficient of variation of raw material heating uniformity at each time point is used as the temperature adaptability of the heating and stirring furnace at each time point.

[0094] The greater the temperature adaptability of the heating and stirring furnace at a single moment, the better the lubricating oil raw material is in the current temperature of the heating and stirring furnace, the better the temperature is suitable, the higher the fluidity and the better the reaction rate.

[0095] The heating and stirring furnace temperature fit scores of all times prior to the current time are arranged chronologically to obtain a heating and stirring furnace temperature fit score sequence. This sequence is then used as input, and the ARIMA prediction algorithm is applied with parameters p, d, q set to (1, 1, 2). The output is the heating and stirring furnace temperature fit score at the current time. Finally, all heating and stirring furnace temperature fit scores are used as input, and cross-validation is employed to determine the optimal segmentation threshold for the heating and stirring furnace temperature fit score. Alternatively, the Otsu thresholding method can be used to obtain the optimal segmentation threshold in another embodiment of this application.

[0096] When the temperature adaptability of the heating and stirring furnace is less than or equal to the optimal segmentation threshold, it is considered that the current temperature of the heating and stirring furnace is no longer suitable for the current lubricating oil raw material heating and stirring process, and needs to be adjusted.

[0097] The intelligent control system for the heating and stirring furnace determines the current temperature suitability of the furnace based on the predicted temperature adaptability and segmentation threshold. When the temperature is suitable, the intelligent control system continuously monitors the furnace. When the temperature is unsuitable, the system outputs a temperature control signal to adjust the heating temperature of the electric heating unit at the bottom of the furnace, thereby achieving a gas boiler temperature control method.

[0098] Based on the same inventive concept as the above method, this application embodiment also provides a gas boiler temperature control system for lubricating oil preparation, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described gas boiler temperature control methods.

[0099] Through the above description of the embodiments in conjunction with the accompanying drawings, those skilled in the art will understand that, for the sake of convenience and brevity, the above division of functional modules is only used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0100] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method of temperature control of a gas boiler, characterized by, The method comprises the following steps: S1, collecting raw material hyperspectral data of the lubricating oil in the heating and stirring furnace at each time point; the raw material hyperspectral data comprises band reflectivity of each pixel corresponding position to all bands; S2, obtaining raw material heating uniformity variation coefficient and raw material heating uniformity at each time point according to the band reflectivity difference and spatial distribution characteristic difference of the pixels in the raw material hyperspectral data at each time point; specifically comprising: S21, obtaining raw material distribution density at each time point according to the band reflectivity difference of the pixels in the raw material hyperspectral data at each time point and the spatial distribution characteristics of the pixels; S22, obtaining raw material stirring fluidity coefficient at each time point according to the band reflectivity difference and spatial distribution characteristic difference of all pixels in the raw material hyperspectral data at adjacent time points; S23, obtaining raw material heating uniformity by combining the raw material stirring fluidity coefficient and the raw material distribution density at each time point; S24, obtaining raw material heating stirring speed at each time point according to the difference in the distribution confusion degree of the band reflectivity of all pixels in the raw material hyperspectral data at adjacent time points; obtaining raw material temperature balance index according to the discrete degree of the change trend of the raw material heating stirring speed at each time point; obtaining raw material heating uniformity variation coefficient according to the numerical distribution of the raw material temperature balance index at all time points; S3, obtaining heating and stirring furnace temperature adaptation degree by combining the raw material heating uniformity and the raw material heating uniformity variation coefficient; controlling the temperature of the heating and stirring furnace according to the numerical distribution of the heating and stirring furnace temperature adaptation degree at all time points before the current time point; The method for obtaining the raw material distribution density at each time point comprises: A sequence composed of the band reflectivity of each pixel corresponding position to all bands is recorded as the spectral sequence of each pixel; the distance between the spectral sequences of any two pixels is taken as the metric distance, and density clustering is performed on all pixels to obtain each raw material clustering cluster; For each raw material clustering cluster, the Euclidean distance between any two pixels is taken as the metric distance, and density clustering is performed on all pixels contained in the raw material clustering cluster to obtain each sub-clustering cluster; For each raw material clustering cluster, the pixel with the minimum horizontal coordinate in each sub-clustering cluster in the raw material clustering cluster is obtained and recorded as the representative pixel of each sub-clustering cluster; for any two sub-clustering clusters, the Euclidean distance between the representative pixels of the two sub-clustering clusters is calculated and recorded as the first distance; the mean value of the first distance of all sub-clustering clusters contained in the raw material clustering cluster is obtained; the ratio of the total number of pixels in the raw material clustering cluster to the mean value is taken as the raw material convergence degree of each raw material clustering cluster; The sum value of the mean values of all raw material clustering clusters is recorded as the total raw material density value; the cumulative sum value of the raw material convergence degrees of all raw material clustering clusters is calculated, and the ratio of the cumulative sum value to the total raw material density value is taken as the raw material distribution density at each time point; The method for obtaining the raw material stirring fluidity coefficient at each time point comprises: The maximum value of the reflectivity of all bands corresponding to the position of each pixel is selected as the representative reflectivity value of each pixel; the Euclidean distance between any two pixels in each raw material clustering cluster is obtained; the average value of all the Euclidean distances in each raw material clustering cluster is obtained; For the raw material hyperspectral data at each time, the information entropy of the representative reflectivity value of all pixels in the raw material hyperspectral data is calculated; the sum value of the average values of all raw material clustering clusters in the raw material hyperspectral data is obtained, which is recorded as the total value of raw material dispersion at each time; According to the numerical distribution difference of the representative reflectivity value of all pixels in the raw material hyperspectral data at adjacent times and the difference of the total value of raw material dispersion, the raw material stirring fluidity coefficient at each time is obtained; The raw material heating uniformity is obtained by combining the raw material stirring fluidity coefficient at each time and the raw material distribution density, including: the ratio of the raw material stirring fluidity coefficient at each time to the raw material distribution density is taken as the raw material heating uniformity at each time; The method for obtaining the raw material temperature balance index at each time specifically includes: A sequence composed of the raw material heating and stirring speeds at a preset number of times before each time is recorded as a trend sequence at each time; The sum value of the average values of all raw material clustering clusters at each time is calculated, which is recorded as the raw material temperature dispersion value at each time; the difference between the raw material temperature dispersion values at each time and the previous time is calculated, which is recorded as a first difference; the variance of the trend sequence at each time is calculated, and the ratio of the first difference to the variance is taken as the raw material temperature balance index at each time.

2. A gas boiler temperature control method as claimed in claim 1, characterized in that, The method for obtaining the raw material stirring fluidity coefficient at each time includes: For each time except the initial time, the difference between the information entropy at each time and the previous time is calculated, which is recorded as an information difference; the difference between the total value of raw material dispersion at each time and the previous time is obtained, which is recorded as a raw material distribution difference; the fusion result of the raw material distribution difference and the information difference at each time and the previous time is recorded as the raw material stirring fluidity coefficient at each time.

3. A gas boiler temperature control method as claimed in claim 1, characterized in that, The method for obtaining the raw material heating and stirring speed at each time includes: The information entropy of the representative reflectivity value of all pixels in the raw material hyperspectral data at each time is calculated, which is recorded as the raw material dispersion value at each time; the ratio of the difference between the raw material dispersion values at each time and the previous time to the preset sampling interval T is recorded as the raw material heating and stirring speed at each time.

4. A gas boiler temperature control method as claimed in claim 1, characterized in that, The method for obtaining the raw material heating uniform change coefficient at each time includes: The trend intensity of the trend sequence at each time is obtained, which is recorded as the raw material mixing trend coefficient at each time; threshold segmentation is performed on the raw material mixing trend coefficients of all times to obtain a mixing judgment threshold; if the raw material mixing trend coefficient at each time is greater than or equal to the mixing judgment threshold, the raw material heating uniform change coefficient at each time is the raw material temperature balance index at each time; otherwise, the raw material heating uniform change coefficient at each time is the reciprocal of the raw material temperature balance index at each time.

5. A gas boiler temperature control method as claimed in claim 1, characterized in that, The heating and stirring furnace temperature adaptation degree is obtained by combining the raw material heating uniformity and the raw material heating uniform change coefficient, and the temperature of the heating and stirring furnace is controlled according to the numerical distribution of the heating and stirring furnace temperature adaptation degree at all times before the current time, including: The product of the raw material heating uniformity at each time and the raw material heating uniformity change coefficient at each time is taken as the heating and stirring furnace temperature adaptation degree at each time; A prediction algorithm is used on the heating and stirring furnace temperature adaptation degrees of all times before the current time to obtain the heating and stirring furnace temperature adaptation degree of the current time; threshold segmentation is performed on the heating and stirring furnace temperature adaptation degrees of all times to obtain an optimal segmentation threshold; If the heating and stirring furnace temperature adaptation degree of the current time is less than or equal to the optimal segmentation threshold, the temperature of the heating and stirring furnace is adjusted.

6. A gas boiler temperature control system for lubricating oil production, comprising a memory, a processor and a computer program stored in the memory and running on the processor, characterized in that, The processor implements the steps of the gas boiler temperature control method according to any one of claims 1-5 when executing the computer program. The processor implements the steps of the gas boiler temperature control method according to any one of claims 1-5 when executing the computer program.

Images