Device and method for identifying coal quality in boiler combustion

Abstract

The present invention relates to a device for identifying coal quality during boiler combustion, comprising an image acquisition device and an information processing device, the image acquisition device and the information processing device being electrically connected. A method for identifying coal quality during boiler combustion comprises the following steps: S1: acquiring and preprocessing a coal quality image; S2: obtaining a connected domain graph based on the coal quality image and calculating the area of ​​each connected domain graph in the coal quality image based on the connected domain graph; S3: obtaining a connected domain area percentage based on the area statistics of each connected domain graph; and S4: comparing the connected domain area percentage with a discrimination threshold to perform graphical discrimination of the coal quality type. The present invention accurately identifies coal quality, providing a necessary prerequisite for boiler combustion optimization.

Description

Technical Field

[0001] The present invention relates to the field of coal quality identification, and in particular to a device and method for identifying coal quality during boiler combustion. Background Art

[0002] Many research institutions and power generation companies have studied how to improve boiler efficiency. Using combustion optimization techniques to maintain boiler operation near optimal combustion conditions is crucial for improving efficiency and reducing pollutant emissions (such as NOx, SO2, CO2, PM10, and PM2.5). However, practical applications present a challenge: frequent fluctuations in the quality of the coal used in boilers can significantly deviate from the overall optimization results if not promptly reflected in parameter configuration. Some studies have attempted to identify coal quality using coal quality image acquisition methods and input them into combustion optimization models. Therefore, achieving this identification has become a pressing issue. Summary of the Invention

[0003] The purpose of the present invention is to overcome the deficiencies of the prior art and provide a device and method for identifying coal quality in boiler combustion, so as to accurately complete coal quality identification and provide the necessary prerequisites for boiler combustion optimization.

[0004] The object of the present invention is achieved through the following technical solutions:

[0005] A device for identifying coal quality during boiler combustion, comprising an image acquisition device and an information processing device, wherein the image acquisition device and the information processing device are electrically connected;

[0006] The image acquisition device acquires coal quality images and sends the coal quality images to the information processing device. After receiving the coal quality images, the information processing device performs the following operations to identify the type of coal quality:

[0007] S1: Preprocessing coal quality image;

[0008] S2: obtaining a connected domain graph according to the coal quality image and calculating the area of ​​each connected domain graph of the coal quality image according to the connected domain graph;

[0009] S3: Obtain the area ratio of connected graphs based on the area statistics of each connected graph;

[0010] S4: Compare the area ratio of the connected graph with the discrimination threshold to perform graphical discrimination of the coal quality type.

[0011] Furthermore, the image acquisition device is fixed by a fixing mechanism, which includes a bracket, a bracket base and a fixed base; the image acquisition device is fixed to the bracket through the fixed base, the bracket is connected to the bracket base, and the bracket base is set on the ground.

[0012] Furthermore, the information processing device is electrically connected to a trigger sensor, and the trigger sensor detects whether the image acquisition device is triggered to perform image acquisition.

[0013] A method for identifying coal quality during boiler combustion, comprising the following steps:

[0014] S1: Collect and preprocess coal quality images;

[0015] S2: obtaining a connected domain graph according to the coal quality image and calculating the area of ​​each connected domain graph of the coal quality image according to the connected domain graph;

[0016] S3: Obtain the area ratio of connected graphs based on the area statistics of each connected graph;

[0017] S4: Compare the area ratio of the connected graph with the discrimination threshold to perform graphical discrimination of the coal quality type.

[0018] Furthermore, step S2 includes the following sub-steps in sequence:

[0019] S201: Standardize the image size of coal quality images;

[0020] S202: Processing the coal quality image using global histogram equalization;

[0021] S203: performing grayscale image binarization processing on the coal quality image;

[0022] S204: performing connectivity enhancement processing on the coal quality image;

[0023] S205: performing graphic segmentation on the coal quality image;

[0024] S206: performing background removal on the coal quality image;

[0025] S207: Perform pixel inversion on the coal quality image to obtain a connected domain graph;

[0026] S208: Calculate the area of ​​each connected graph of the coal quality image based on the connected domain graph.

[0027] Furthermore, step S3 includes the following sub-steps:

[0028] S301: performing error elimination processing on the area of ​​each connected graph;

[0029] S302: Counting the area of ​​each connected graph according to the type of connected graph to obtain a connected graph area ratio.

[0030] Furthermore, the discrimination threshold is obtained by a heuristic discrimination threshold calculation method based on small sample statistics; the heuristic discrimination threshold calculation method based on small sample statistics comprises the following steps:

[0031] (1) Collect coal quality images for each type of coal multiple times and obtain the corresponding connected graph area ratio;

[0032] (2) setting an estimated discrimination threshold for each coal quality type according to the corresponding connected graph area ratio;

[0033] (3) defining an optional value range for the estimated discrimination threshold parameter based on the characteristics of the discrimination threshold and the relationship between the discrimination threshold and the corresponding connected graph area ratio;

[0034] (4) determining a parameter fusion value according to the optional value range, and calculating a discrimination threshold estimation value according to the parameter fusion value and the corresponding connected graph area ratio;

[0035] (5) The discrimination threshold is obtained by heuristic iterative calculation using the estimated value of the discrimination threshold.

[0036] Furthermore, the estimated discrimination threshold obtained when the coal quality type is the i-th coal is set as:

[0037]

[0038] The optional value range is and

[0039]

[0040] Among them, Δ j is the parameter of the estimated discrimination threshold, th j is the estimated discrimination threshold for the jth acquisition, is the area ratio of the corresponding connected graph collected for the jth time, j represents the number of times each type of coal is collected, the maximum is M times, i represents the type of connected graph, and there are N types in total.

[0041] Furthermore, the area ratio of the connected graph obtained in step S3 is corrected by a dynamic correction method based on average grayscale, and the dynamic correction method based on average grayscale includes the following steps:

[0042] 1) Calculate the average grayscale of the original image of the coal quality image;

[0043] 2) calculating the grayscale mean of multiple coal qualities based on the coal quality images collected multiple times for each type of coal quality in step (1) of the heuristic discrimination threshold calculation method based on small sample statistics;

[0044] 3) Calculate the correction factor based on the average grayscale of the original image of the coal quality image and the grayscale mean of various coal qualities;

[0045] 4) Correct the area ratio of the connected graph according to the correction factor.

[0046] Furthermore, the step S4 further includes the following steps:

[0047] S5: performing grayscale discrimination of the original image of the coal quality image based on the variable scale grid segmentation image block to obtain the variable scale grid grayscale discrimination of the coal quality type;

[0048] S6: Combine the results of the graphical discrimination of the coal quality type and the variable-scale grid grayscale discrimination of the coal quality type to obtain the final result.

[0049] The beneficial effects of the present invention are:

[0050] Accurately identify coal quality and provide the necessary prerequisites for boiler combustion optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 This is a schematic diagram of the mechanical structure of a device for identifying coal quality during boiler combustion;

[0052] Figure 2 Schematic diagram for bracket height calculation;

[0053] Figure 3 Schematic diagram of the structure of the bracket base;

[0054] Figure 4 is a plan view of the bracket base;

[0055] Figure 5 It is a plan view of the top surface of the underground pile;

[0056] Figure 6 It is a structural diagram of the fixed base;

[0057] Figure 7 is a plan view of the first connection surface;

[0058] Figure 8 is a plan view of the second connection surface;

[0059] Figure 9 is a top view of the sunshade;

[0060] Figure 10 This is a schematic diagram of the electrical structure of a device for identifying coal quality during boiler combustion;

[0061] Figure 11 These are three types of coal quality pictures collected under the same conditions;

[0062] Figure 12 This is the image obtained after processing the various sub-steps of step S2 (according to the sub-steps from left to right in the first row and then from left to right in the second row).

[0063] In the figure, 1-bracket, 2-bracket base, 3-fixed base, 4-light shield, 5-industrial black and white camera, 6-fluxgate sensor, 7-fill light, 8-top surface of underground pile, 9-first nut, 10-first screw, 11-first connecting surface, 12-second connecting surface, 13-connecting rod welding point, 14-collective harness hole, 15-fill light harness hole, 16-fluxgate sensor harness hole, 17-camera harness hole, 18-fixed connecting rod, 19-camera fixing hole. DETAILED DESCRIPTION

[0064] The following describes the embodiments of the present invention through specific examples. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments. The details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other unless they conflict.

[0065] It should be noted that the illustrations provided in the following embodiments are merely schematic illustrations of the basic concept of the present invention. Therefore, the illustrations only show components related to the present invention and are not drawn according to the number, shape, and size of components in actual implementation. In actual implementation, the type, quantity, and proportion of each component may be changed arbitrarily, and the component layout may also be more complex.

[0066] Example 1:

[0067] like Figures 1 to 12 As shown, a device for identifying coal quality during boiler combustion includes an image acquisition device and an information processing device, wherein the image acquisition device and the information processing device are electrically connected;

[0068] The image acquisition device acquires coal quality images and sends the coal quality images to the information processing device. After receiving the coal quality images, the information processing device performs the following operations to identify the type of coal quality:

[0069] S1: Preprocessing coal quality image;

[0070] S2: obtaining a connected domain graph according to the coal quality image and calculating the area of ​​each connected domain graph of the coal quality image according to the connected domain graph;

[0071] S3: Obtain the area ratio of connected graphs based on the area statistics of each connected graph;

[0072] S4: Compare the area ratio of the connected graph with the discrimination threshold to perform graphical discrimination of the coal quality type.

[0073] The step S2 includes the following sub-steps in sequence:

[0074] S201: Standardize the image size of coal quality images;

[0075] S202: Processing the coal quality image using global histogram equalization;

[0076] S203: performing grayscale image binarization processing on the coal quality image;

[0077] S204: performing connectivity enhancement processing on the coal quality image;

[0078] S205: performing graphic segmentation on the coal quality image;

[0079] S206: performing background removal on the coal quality image;

[0080] S207: Perform pixel inversion on the coal quality image to obtain a connected domain graph;

[0081] S208: Calculate the area of ​​each connected graph of the coal quality image based on the connected domain graph.

[0082] The step S3 includes the following sub-steps:

[0083] S301: performing error elimination processing on the area of ​​each connected graph;

[0084] S302: Counting the area of ​​each connected graph according to the type of connected graph to obtain a connected graph area ratio.

[0085] The discrimination threshold is obtained by a heuristic discrimination threshold calculation method based on small sample statistics; the heuristic discrimination threshold calculation method based on small sample statistics comprises the following steps:

[0086] (1) Collect coal quality images for each type of coal multiple times and obtain the corresponding connected graph area ratio;

[0087] (2) setting an estimated discrimination threshold for each coal quality type according to the corresponding connected graph area ratio;

[0088] (3) defining an optional value range for the estimated discrimination threshold parameter based on the characteristics of the discrimination threshold and the relationship between the discrimination threshold and the corresponding connected graph area ratio;

[0089] (4) determining a parameter fusion value according to the optional value range, and calculating a discrimination threshold estimation value according to the parameter fusion value and the corresponding connected graph area ratio;

[0090] (5) The discrimination threshold is obtained by heuristic iterative calculation using the estimated value of the discrimination threshold.

[0091] The estimated discrimination threshold obtained when the coal quality type is the i-th coal is set as:

[0092]

[0093] The optional value range is and

[0094]

[0095] Among them, Δ j is the parameter of the estimated discrimination threshold, th j is the estimated discrimination threshold for the jth acquisition, is the area ratio of the corresponding connected graph collected for the jth time, j represents the number of times each type of coal is collected, the maximum is M times, i represents the type of connected graph, and there are N types in total.

[0096] The area ratio of the connected graph obtained in step S3 is corrected by a dynamic correction method based on average grayscale, and the dynamic correction method based on average grayscale includes the following steps:

[0097] 1) Calculate the average grayscale of the original image of the coal quality image;

[0098] 2) calculating the grayscale mean of multiple coal qualities based on the coal quality images collected multiple times for each type of coal quality in step (1) of the heuristic discrimination threshold calculation method based on small sample statistics;

[0099] 3) Calculate the correction factor based on the average grayscale of the original image of the coal quality image and the grayscale mean of various coal qualities;

[0100] 4) Correct the area ratio of the connected graph according to the correction factor.

[0101] After step S4, the following steps are also included:

[0102] S5: performing grayscale discrimination of the original image of the coal quality image based on the variable scale grid segmentation image block to obtain the variable scale grid grayscale discrimination of the coal quality type;

[0103] S6: Combine the results of the graphical discrimination of the coal quality type and the variable-scale grid grayscale discrimination of the coal quality type to obtain the final result.

[0104] The image acquisition device is a camera.

[0105] The image acquisition device is fixedly installed in the coal yard to acquire images of coal quality.

[0106] The image acquisition device is arranged at the weighing place of the coal transport vehicle through the bracket 1, that is, when the coal transport vehicle is weighed at the weighing place, the image acquisition device is used to vertically shoot the coal in the vehicle bucket from a high position.

[0107] This is because the image acquisition equipment requires stable power supply and network communication, as well as the operating environment of the coal yard in the power plant.

[0108] The image acquisition device is fixed by a fixing mechanism, which includes a bracket 1, a bracket base 2 and a fixed base 3; the image acquisition device is fixed to the bracket 1 through the fixed base 3, the bracket 1 is connected to the bracket base 2, and the bracket base 2 is set on the ground.

[0109] The bracket 1 is connected to the bracket base 2 as the supporting structure of the image acquisition device, and is formed by welding multiple sections of steel pipes. Figure 1 The structure shown is connected inside the pipe for line layout.

[0110] The image acquisition device is installed at the weighing position of the coal transport vehicle through the bracket 1. The image acquisition device is fixed to the bracket 1 through a fixing device. The shooting part of the image acquisition device is positioned at the center of the coal transport vehicle's cargo box. The height of the bracket 1 is calculated according to the following formula to meet the use requirements and the minimum requirements of the shooting part coverage:

[0111] h=h1+h2+h3

[0112] h1=d / (2tan(a / 2))

[0113] Where h represents the height of the bracket 1, h1 represents the required height for shooting, h2 represents the height of the coal truck, h3 represents the height of the image acquisition device and the fixing device, d represents the width of the truck bed, and a represents the viewing angle of the camera unit.

[0114] The bracket base 2 is connected to the ground and serves as a fixed ground (metal) structure of the bracket 1. The specific structure is as follows: Figure 3 As shown. The underground pile is fixed by pouring cement, and a ground connection is reserved for the top surface 8 (metal surface) of the underground pile. The bracket base 2 and the top surface 8 of the underground pile are fastened by four bolts (including a first nut 9 and a first screw 10);

[0115] The fixed base 3, serving as the fixed connection surface for the image acquisition device, is connected to the bracket 1. The image acquisition device is secured to the fixed base 3 using screws, strong glue, or other fixing methods. The fixed base 3 includes a first connection surface 11 and a second connection surface 12, which are welded together by four fixed connecting rods 18 (metal rods). Holes are drilled between the first connection surface 11 and the second connection surface 12 for routing electrical and signal lines. The first connection surface 11 is welded to the bracket 1.

[0116] The first connecting surface 11 and the second connecting surface 12 are both metal surfaces.

[0117] A fixed connecting rod welding point 13 is provided on each of the first connecting surface 11 and the second connecting surface 12 for welding to the fixed connecting rod 18 .

[0118] A wiring harness hole 14 is provided on the first connecting surface 11, and a wiring harness hole 15 for the fill light 7, a wiring harness hole 16 for the fluxgate sensor 6 and a wiring harness hole 17 for the camera are provided on the second connecting surface 12, which are used for routing the fill light 7, the fluxgate sensor 6 and the camera respectively. The routing of the three passes through the wiring harness hole 14 on the first connecting surface 11 and enters the steel pipe connected to the pipe provided on the bracket 1.

[0119] The second connection surface 12 is further provided with a camera fixing hole 19 for fixing the camera by means of bolts.

[0120] A light shield 4 is provided on the top of the bracket 1 .

[0121] The lens hood 4 reduces the influence of light on the photograph to a certain extent.

[0122] A fill light 7 is provided below the light shield 4 , and the fill light 7 cooperates with the light shield 4 to provide a relatively stable illumination.

[0123] The camera is an industrial black-and-white camera 5. It is responsible for taking industrial-grade black-and-white photos. Since the subsequent image processing method also only requires grayscale images, the black-and-white camera can provide images with richer details under this requirement.

[0124] The information processing device is electrically connected to a trigger sensor, and the trigger sensor detects whether to trigger the image acquisition device to perform image acquisition.

[0125] The trigger sensor is a fluxgate sensor 6 .

[0126] The information processing device includes a host computer and a slave computer.

[0127] The host computer receives the signal of the trigger sensor and determines whether the image acquisition device performs image acquisition; when the judgment result of the host computer is that the image acquisition device performs image acquisition, the host computer controls the image acquisition device and the fill light 7 to work; the host computer is provided with a temporary storage device, which stores the coal quality images that need to be temporarily stored.

[0128] When the coal transport vehicle approaches the image acquisition device, the geomagnetic signal collected by the fluxgate sensor 6 changes. The host computer receives the signal sent by the fluxgate sensor 6 and detects that the change gradient of the geomagnetic signal collected by the fluxgate sensor 6 is 0 (that is, the coal transport vehicle is stopped). The host computer determines that the image acquisition device is performing image acquisition and sends a signal to the image acquisition device to trigger shooting.

[0129] When shooting, the camera must be perpendicular to the ground. And to ensure that the light source of the fill light 7 is stable after it is turned on, a delayed shooting setting is required.

[0130] In order to ensure image quality and avoid instability caused by improper focus during shooting, multiple shots are required, and the lower computer makes the final selection.

[0131] At the request of the lower computer, the upper computer sends the coal quality image stored in the temporary storage device to the lower computer, and completes the identification of the coal quality type based on the coal quality image.

[0132] A method for identifying coal quality during boiler combustion, comprising the following steps:

[0133] S1: Collect and preprocess coal quality images;

[0134] S2: obtaining a connected domain graph according to the coal quality image and calculating the area of ​​each connected domain graph of the coal quality image according to the connected domain graph;

[0135] The area of ​​each connected graph is obtained according to the connected domain graph of the coal quality image, and the area of ​​each connected graph is statistically analyzed to obtain the approximate size and uniformity of the coal block.

[0136] This embodiment uses a graphics-based method to determine the size and uniformity of coal particles.

[0137] Small and uniform coal particles indicate good coal quality, while large and uneven coal particles indicate poor coal quality. Therefore, the area of ​​each connected graph is calculated to represent the coal quality.

[0138] Step S2 includes the following sub-steps in sequence:

[0139] S201: Standardize the image size of coal quality images;

[0140] S202: Processing the coal quality image using global histogram equalization;

[0141] S203: performing grayscale image binarization processing on the coal quality image;

[0142] S204: performing connectivity enhancement processing on the coal quality image;

[0143] Using morphological methods, the coal image is eroded and then expanded for two rounds; the coal image is then expanded for three rounds to increase the connectivity of the image; and median filtering is performed;

[0144] S205: performing graphic segmentation on the coal quality image;

[0145] S205 uses the watershed method to perform image segmentation and find the foreground area.

[0146] S206: performing background removal on the coal quality image;

[0147] Find the unknown area, fill it with 0, and obtain the marked area map;

[0148] S207: Perform pixel inversion on the coal quality image to obtain a connected domain graph;

[0149] S208: Calculating the area of ​​each connected graph of the coal quality image according to the connected domain graph; merging the connected domain graphs to obtain a coal block contour map.

[0150] S3: Obtain the area ratio of connected graphs based on the area statistics of each connected graph;

[0151] Since there are generally not many types of coal qualities, statistical characteristics can be set according to the actual coal types used.

[0152] S301: performing error elimination processing on the area of ​​each connected graph;

[0153] Taking into account the effects of particle shadows and errors, the connected graph is filtered out if the area is within the filter area range, which is less than 5,000 pixels or greater than 50,000 pixels.

[0154] S302: Counting the area of ​​each connected graph according to the type of connected graph to obtain a connected graph area ratio;

[0155] Connectivity graph types are categorized by the size of their connected area, which indicates coal quality. Different coal qualities have different connected area sizes. For example, connectivity graph types include small connected graphs (with a connected area of ​​less than 10,000 pixels), medium connected graphs (with a connected area of ​​10,000 to 30,000 pixels), and large connected graphs (with a connected area of ​​more than 30,000 pixels).

[0156] like Figure 11 In the middle, from left to right, they belong to the small connected graph, the medium connected graph, and the large connected graph, which are the coal quality images of No. 1 coal, No. 2 coal, and No. 3 coal respectively.

[0157] S4: Compare the area ratio of the connected graph with the discrimination threshold to perform graphical discrimination of coal quality;

[0158] For example, if the area of ​​the small connected graph accounts for more than 90% (the discrimination threshold of the small connected graph is 0.9), it can be considered as No. 1 coal; if the area of ​​the medium connected graph accounts for 20% (the discrimination threshold of the medium connected graph is 0.2), it can be considered as No. 2 coal; if the area of ​​the large connected graph accounts for 20% (the discrimination threshold of the large connected graph is 0.2), it can be considered as No. 3 coal.

[0159] Through steps S1-S4, the discrimination based on graphics is realized, and the coal quality is accurately identified, which provides the necessary prerequisite for boiler combustion optimization.

[0160] The area ratio of the connected graph is represented by the vector s=[s1,s2,...,s N ]express;

[0161] The threshold is determined by the threshold vector th=[th1,th2,...,th N ]express;

[0162] The discrimination result of the comparative discrimination is represented by the vector α=[α1,α2,...,α N ] indicates; among them

[0163]

[0164] i represents the type of connected graph, i.e. the type of coal quality, which has N types in total;

[0165] The discrimination threshold is obtained by a heuristic discrimination threshold calculation method based on small sample statistics;

[0166] Discrimination threshold th=[th1,th2,...,th N ] can be obtained through a small amount of experimental statistical approximation.

[0167] The heuristic discrimination threshold calculation method based on small sample statistics includes the following steps:

[0168] (1) Collect coal quality images for each type of coal multiple times and obtain the corresponding connected graph area ratio;

[0169] This step collects coal quality images and obtains the corresponding connected graph area ratio by executing steps S1, S2, and S3 of a method for identifying coal quality in boiler combustion.

[0170] The corresponding connected graph area ratio is Where j represents the number of times each type of coal is collected, with a maximum of M times.

[0171] The discriminant result vector should be:

[0172] (2) setting an estimated discrimination threshold for each coal quality type according to the corresponding connected graph area ratio;

[0173] The estimated discrimination threshold obtained when the coal quality type is the i-th coal is set as:

[0174]

[0175] The above formula is the estimated discrimination threshold obtained when the coal quality type is type i coal;

[0176] The estimated discrimination threshold is set according to the above formula because the discrimination threshold needs to meet all M situations. The coal quality type of the collected image is judged as the type of connected graph corresponding to the i-th type of coal, which means that the threshold th i j<s i j , the remaining thresholds Therefore, the estimated discrimination threshold is ±Δ j Indicates that Δ j This is the parameter of the estimated discrimination threshold.

[0177] (3) defining an optional value range for the estimated discrimination threshold parameter based on the characteristics of the discrimination threshold and the relationship between the discrimination threshold and the corresponding connected graph area ratio;

[0178] First of all i j >0, then in formula (2)

[0179] That is

[0180] make

[0181] Then it satisfies

[0182] Assume that the true decision threshold exists in th j In

[0183] and

[0184] From the above formula, we can deduce

[0185]

[0186]

[0187] make

[0188] Then it satisfies

[0189] Will and Combining the two inequalities we get:

[0190]

[0191] The optional value range is and

[0192] Among them, Δ j is the parameter of the estimated discrimination threshold, th j is the estimated discrimination threshold for the jth acquisition, is the area ratio of the corresponding connected graph collected for the jth time, j represents the number of times each type of coal is collected, the maximum is M times, i represents the type of connected graph, and there are N types in total.

[0193] (4) determining a parameter fusion value according to the optional value range, and calculating a discrimination threshold estimation value according to the parameter fusion value and the corresponding connected graph area ratio;

[0194] Assume that there are p Δ in the optional value range, and mark it as set P,

[0195] In fact, the smaller the Δ in the optional value range is, the larger the area of ​​the connected graph of the group is. j The closer it is to the discrimination threshold th, the average value of the lower limit is taken as the parameter fusion value in this embodiment:

[0196] The estimated value of the discrimination threshold is That is, use any set of corresponding connected graph area proportions that satisfy the inequality conditions Vector overlay parameter blend values.

[0197] The parameter fusion value can also be determined in other ways, such as mean, median, or even random selection.

[0198] (5) The discrimination threshold is obtained by heuristic iterative calculation using the estimated value of the discrimination threshold.

[0199] Using the discrimination threshold estimate Replace the corresponding connected graph area ratio selected in the calculation of the discrimination threshold estimate Vector, repeat steps (2), (3), (4) to obtain the new parameter fusion value Δ * and the estimated value of the discrimination threshold Until the parameter fusion value Δ * No longer changes, and the estimated value of the discrimination threshold calculated this time is This is the final judgment threshold.

[0200] The area ratio of the connected graph obtained in step S3 is corrected by a dynamic correction method based on average grayscale;

[0201] Since the method based on graphics analysis is based on the difference in coal block size, the influence of coal composition is not considered. However, coal composition will cause changes in the grayscale of the coal image (for example, high moisture content in the coal will result in low grayscale values). When performing graphic segmentation, the inaccurate edges after image preprocessing will lead to "over-connectivity", that is, the area of ​​the connected graph is too large. Therefore, when the area of ​​the connected graph is counted, deviations will occur due to changes in the average grayscale of the image. This embodiment proposes a method based on dynamic correction of the average grayscale, which uses the average grayscale of the image to dynamically correct statistical characteristics.

[0202] The dynamic correction method based on average grayscale comprises the following steps:

[0203] 1) Calculate the average grayscale g* of the original image of the coal quality image.

[0204] 2) calculating the grayscale mean of multiple coal qualities based on the coal quality images collected multiple times for each type of coal quality in step (1) of the heuristic discrimination threshold calculation method based on small sample statistics;

[0205] The image used for reference in determining the proportion of statistical characteristics is the coal quality image obtained by collecting multiple times of each type of coal in step (1) of the heuristic discrimination threshold calculation method based on small sample statistics, and the grayscale mean of each type of coal quality image is calculated according to the coal quality type, which are g1, g2, ..., g N ; Then calculate the grayscale mean of various coal qualities, that is, the average grayscale of N kinds of coal qualities, that is, g avg =(g1+g2+...+g N ) / N.

[0206] 3) The correction factor is calculated based on the average grayscale of the original coal quality image and the grayscale mean of various coal qualities. The correction factor is f = g* / g avg .

[0207] 4) Correct the area ratio of the connected graph according to the correction factor;

[0208] The area ratio of the corrected connected graph is the area ratio of the original connected graph multiplied by the correction factor f.

[0209] That is, s′=f·[s1,s2,...,s N ]

[0210] Finally determine whether it meets the proportion characteristics of a certain coal quality.

[0211] After step S4, the following steps are also included:

[0212] S5: performing grayscale discrimination of the original image of the coal quality image based on the variable scale grid segmentation image block to obtain the variable scale grid grayscale discrimination of the coal quality type;

[0213] The original image of the coal quality image is subjected to variable scale gridding processing, wherein the number of scales of the variable scale gridding processing is determined according to the type of coal quality, the grayscale of each image block at each scale is calculated, and the variable scale grid grayscale discrimination of the coal quality type is performed based on the grayscale of each image block;

[0214] The original image of the coal quality image is Gridding, where U is the number of scales, generally let U = N, R = [10, 20, 30, ..., 10 * U], that is, the number of scales is selected according to the number of coal types. Calculate R at each scale separately o ×R o The average grayscale of each image block is sorted by the 1 / 0 2 Uniform sampling (ensuring the same number of comparisons for each scale) is performed to compare the grayscale values. The larger the grayscale value, the more points it will get. The segmentation scale with the lowest score will be eliminated. Similarly, the comparison between the top 20% to 10% will be performed again. And so on. The final winning scale o is used to determine the type of coal quality.

[0215] S6: combining the results of the coal quality classification graphical discrimination and the coal quality classification variable scale grid grayscale discrimination to obtain the final result;

[0216] Combining the two discrimination results of coal quality type graphics discrimination and coal quality type variable scale grid grayscale discrimination can obtain a more robust discrimination result.

[0217] If the two discrimination results are consistent, the identification result is clear. If the two discrimination results are inconsistent, in order to ensure more reliable parameter control in the future, the calculation is based on the medium quality of coal, that is, it is identified as intermediate grade coal at this time.

[0218] When the graphical discrimination of coal quality type is unclear or not obvious, the variable scale grid grayscale discrimination of coal quality type can also be used.

[0219] A device for identifying coal quality during boiler combustion uses a method for identifying coal quality during boiler combustion.

[0220] The above-described embodiments merely represent specific implementations of the present invention. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, and all such variations and improvements fall within the scope of protection of the present invention.

Claims

1. A device for identifying coal quality during boiler combustion, characterized by: It includes an image acquisition device and an information processing device, wherein the image acquisition device is electrically connected to the information processing device; The image acquisition device acquires coal quality images and sends the coal quality images to the information processing device. After receiving the coal quality images, the information processing device performs the following operations to identify the type of coal quality: S1: Preprocessing coal quality image; S2: obtaining a connected domain graph according to the coal quality image and calculating the area of ​​each connected domain graph of the coal quality image according to the connected domain graph; S3: Obtain the area ratio of connected graphs based on the area statistics of each connected graph; S4: Compare the area ratio of the connected graph with the discrimination threshold to perform graphical discrimination of coal quality; S5: performing grayscale discrimination of the original image of the coal quality image based on the variable scale grid segmentation image block to obtain the variable scale grid grayscale discrimination of the coal quality type; S6: combining the results of the coal quality classification graphical discrimination and the coal quality classification variable scale grid grayscale discrimination to obtain the final result; The discrimination threshold is obtained by a heuristic discrimination threshold calculation method based on small sample statistics; the heuristic discrimination threshold calculation method based on small sample statistics comprises the following steps: (1) Collect coal quality images for each type of coal multiple times and obtain the corresponding connected graph area ratio; (2) setting an estimated discrimination threshold for each coal quality type according to the corresponding connected graph area ratio; (3) defining an optional value range for the estimated discrimination threshold parameter based on the characteristics of the discrimination threshold and the relationship between the discrimination threshold and the corresponding connected graph area ratio; (4) determining a parameter fusion value according to the optional value range, and calculating a discrimination threshold estimation value according to the parameter fusion value and the corresponding connected graph area ratio; (5) The discrimination threshold is obtained by heuristic iterative calculation using the estimated value of the discrimination threshold.

2. The device for identifying coal quality during boiler combustion according to claim 1, characterized in that: The image acquisition device is fixed by a fixing mechanism, which includes a bracket, a bracket base and a fixed base; the image acquisition device is fixed on the bracket through the fixed base, the bracket is connected to the bracket base, and the bracket base is set on the ground.

3. The device for identifying coal quality during boiler combustion according to claim 1, characterized in that: The information processing device is electrically connected to a trigger sensor, and the trigger sensor detects whether to trigger the image acquisition device to perform image acquisition.

4. A method for identifying coal quality during boiler combustion, characterized in that: The following steps are involved: S1: Collect and preprocess coal quality images; S2: obtaining a connected domain graph according to the coal quality image and calculating the area of ​​each connected domain graph of the coal quality image according to the connected domain graph; S3: Obtain the area ratio of connected graphs based on the area statistics of each connected graph; S4: Compare the area ratio of the connected graph with the discrimination threshold to perform graphical discrimination of coal quality; S5: performing grayscale discrimination of the original image of the coal quality image based on the variable scale grid segmentation image block to obtain the variable scale grid grayscale discrimination of the coal quality type; S6: combining the results of the coal quality classification graphical discrimination and the coal quality classification variable scale grid grayscale discrimination to obtain the final result; The discrimination threshold is obtained by a heuristic discrimination threshold calculation method based on small sample statistics; the heuristic discrimination threshold calculation method based on small sample statistics comprises the following steps: (1) Collect coal quality images for each type of coal multiple times and obtain the corresponding connected graph area ratio; (2) setting an estimated discrimination threshold for each coal quality type according to the corresponding connected graph area ratio; (3) defining an optional value range for the estimated discrimination threshold parameter based on the characteristics of the discrimination threshold and the relationship between the discrimination threshold and the corresponding connected graph area ratio; (4) determining a parameter fusion value according to the optional value range, and calculating a discrimination threshold estimation value according to the parameter fusion value and the corresponding connected graph area ratio; (5) The discrimination threshold is obtained by heuristic iterative calculation using the estimated value of the discrimination threshold.

5. The method for identifying coal quality during boiler combustion according to claim 4, characterized in that: The step S2 includes the following sub-steps in sequence: S201: Standardize the image size of coal quality images; S202: Processing the coal quality image using global histogram equalization; S203: performing grayscale image binarization processing on the coal quality image; S204: performing connectivity enhancement processing on the coal quality image; S205: performing graphic segmentation on the coal quality image; S206: performing background removal on the coal quality image; S207: Perform pixel inversion on the coal quality image to obtain a connected domain graph; S208: Calculate the area of ​​each connected graph of the coal quality image based on the connected domain graph.

6. The method for identifying coal quality during boiler combustion according to claim 4, characterized in that: The step S3 includes the following sub-steps: S301: performing error elimination processing on the area of ​​each connected graph; S302: Counting the area of ​​each connected graph according to the type of connected graph to obtain a connected graph area ratio.

7. The method for identifying coal quality during boiler combustion according to claim 4, characterized in that: The estimated discrimination threshold obtained when the coal quality type is the i-th coal is set as: The optional value range is k∈[1,N],j∈[1,M] and Δ j >0, Among them, Δ j is the parameter of the estimated discrimination threshold, th j is the estimated discrimination threshold for the jth acquisition, is the area ratio of the corresponding connected graph collected for the jth time, j represents the number of times each type of coal is collected, the maximum is M times, i represents the type of connected graph, and there are N types in total.

8. A method for identifying coal quality during boiler combustion according to claim 4 or 7, characterized in that: The area ratio of the connected graph obtained in step S3 is corrected by a dynamic correction method based on average grayscale, and the dynamic correction method based on average grayscale includes the following steps: 1) Calculate the average grayscale of the original image of the coal quality image; 2) calculating the grayscale mean of multiple coal qualities based on the coal quality images collected multiple times for each type of coal quality in step (1) of the heuristic discrimination threshold calculation method based on small sample statistics; 3) Calculate the correction factor based on the average grayscale of the original image of the coal quality image and the grayscale mean of various coal qualities; 4) Correct the area ratio of the connected graph according to the correction factor.

Images