Fly ash carbon content on-line soft measurement device and method based on multi-scale coupling

Abstract

The invention discloses a fly ash carbon content online soft measurement device and method based on multi-scale coupling. The device comprises a sampling pipe, a flue, a rectangular sample pool, a sample valve, a sample gate, a transparent conveyor belt, a microscope, a light source, an infrared receiver, a fixed-frequency fan, a fly ash collecting pipe, a fly ash collecting bag, an infrared transmitter, a background baffle and an operation module. Wherein the microscope is used for carrying out continuous microscopic image acquisition on a fly ash sample; the infrared transmitter and the infrared receiver are used for continuously collecting light transmission data of the fly ash sample; the fly ash collecting pipe and the fly ash collecting bag are used for recording the fly ash weight ratio of the two outlets; and the operation module is in communication connection with each acquisition device, and is used for carrying out dynamic analysis on the continuously acquired microscopic images and the fly ash physical property parameters and outputting the fly ash carbon content. According to the device, the carbon content of the fly ash is measured in a manner of combining continuous microscopic image characteristics and physical property characteristics, so that the speed and the accuracy of measuring the carbon content of the fly ash are improved.

Description

Technical Field

[0001] The present application relates to the technical field of fly ash carbon content detection, and in particular to an online soft measurement device and method for fly ash carbon content based on multi-scale coupling. Background Art

[0002] At present, in modern industrial production, the operating efficiency and environmental performance of pulverized coal boilers in heat source plants are receiving more and more attention. Among them, the carbon content of fly ash is one of the important indicators that affect and evaluate the efficiency of boilers. As the residue after coal combustion, the carbon content of fly ash directly reflects the degree of coal combustion. For every 1% decrease in fly ash carbon content, the boiler's combustion efficiency can increase by 0.51%. Excessive fly ash carbon content will not only reduce the service life of the boiler, but also mean a waste of energy and possible pollution to the environment. Therefore, the detection of fly ash carbon content is an important part of the work process of pulverized coal boilers.

[0003] In the related art, the traditional method of measuring the carbon content of fly ash usually relies on offline sampling and laboratory analysis, for example, the carbon content is measured through the loss on ignition LOI test, but this measurement method has the disadvantages of poor timeliness, cumbersome operation and inability to reflect the boiler operation status in real time. In order to meet the demand for real-time measurement of the carbon content of fly ash, some methods that can directly and quickly measure the carbon content of fly ash have begun to be adopted in the related art, such as microwave method, optical reflection method or radiation method. However, although these measurement methods can improve the measurement speed, the measurement accuracy is poor and the measurement results may be biased.

[0004] Therefore, how to quickly measure the carbon content of fly ash while ensuring measurement accuracy has become a problem that needs to be solved urgently. Summary of the invention

[0005] The purpose of this application is to solve one of the above-mentioned technical problems at least to a certain extent.

[0006] To this end, the first purpose of the present application is to propose an online soft measurement device for fly ash carbon content based on multi-scale coupling. The device can measure the fly ash carbon content by combining continuous microscopic image features and physical property features, thereby improving the speed and accuracy of fly ash carbon content measurement.

[0007] The second objective of this application is to propose an online soft measurement method for fly ash carbon content based on multi-scale coupling.

[0008] A third object of the present application is to provide a non-transitory computer-readable storage medium.

[0009] To achieve the above-mentioned purpose, the first embodiment of the present application proposes an online soft measurement device for carbon content in fly ash based on multi-scale coupling, the device comprising: a sampling tube, a flue, a rectangular sample pool, a sample valve, a sample gate, a transparent conveyor belt, a microscope, a light source, an infrared receiver, a fixed-frequency fan, a fly ash collection tube, a fly ash collection bag, an infrared transmitter, a background baffle and a computing module; wherein,

[0010] The sampling tube is installed in the downstream area of ​​the flue, the sampling tube is used to sample the fly ash in the flue, and the rectangular sample pool is used to store the collected fly ash;

[0011] The sample valve is equipped with a gravity sensor. When the weight of the fly ash falling from the rectangular sample pool reaches a preset weight, the sample valve opens to transport the fly ash to the transparent conveyor belt below; the transparent conveyor belt is used to drive the fly ash to move, and the sample gate is used to cut the fly ash passing through into fly ash samples with uniform thickness;

[0012] The microscope is used to continuously collect microscopic images of the fly ash sample with the assistance of the light source and the background baffle; the infrared receiver is located behind the microscope, and the infrared transmitter and the infrared receiver are used to continuously collect light transmittance data of the fly ash sample; the fly ash collection pipe and the fly ash collection bag are used to record the fly ash weight ratio of the two outlets;

[0013] The operation module is in communication connection with each acquisition device, and is used for dynamically analyzing the continuously acquired microscopic images and fly ash physical property parameters to output the carbon content of the fly ash.

[0014] In addition, the online soft measurement device for carbon content in fly ash based on multi-scale coupling in the embodiment of the present application also has the following additional technical features:

[0015] Optionally, in some embodiments, the device also includes: a sampling valve, an exhaust pipe, a cyclone separator and a separator outlet; wherein the output end of the sampling tube is connected to the input end of the sampling valve, the output end of the sampling valve is connected to the input end of the cyclone separator, and the cyclone separator is used to separate the gas and solid in the collected fly ash; the separator outlet is connected between the cyclone separator and the rectangular sample pool, and the separated solid fly ash passes through the separator outlet under the action of centrifugal force and enters the rectangular sample pool.

[0016] Optionally, in some embodiments, the device further includes: a cleaning baffle, the cleaning baffle being located below the rectangular sample pool, the bottom of the cleaning baffle being made of a brush material; the cleaning baffle being used to block the background and clean impurities on the conveyor belt to be fed into the rectangular sample pool.

[0017] Optionally, in some embodiments, the device further includes: a ash cleaner, which is located below the rectangular sample pool and is used in combination with the cleaning baffle to clean the conveyor belt to be introduced into the rectangular sample pool; and a conveyor belt roller, which is used to drive the transparent conveyor belt to operate.

[0018] Optionally, in some embodiments, the microscope includes: a data transmission module, which is used to transmit the collected microscopic image to the storage device of the computing module in real time; the infrared receiver is also used to transmit the collected light-transmitting data to the storage device of the computing module.

[0019] Optionally, in some embodiments, the computing module is equipped with a trained deep learning network model, and the deep learning network model is used to predict the fly ash carbon content corresponding to the current time period based on the real-time collected microscopic image time series, infrared transmittance time series, the fly ash weight ratio and boiler operation data.

[0020] To achieve the above object, the second embodiment of the present invention proposes an online soft measurement method for carbon content in fly ash based on multi-scale coupling, which is applied to the online soft measurement device for carbon content in fly ash based on multi-scale coupling of the first aspect. The method comprises:

[0021] Fly ash is collected in the flue through a sampling tube, and the collected fly ash is concentrated by a cyclone separator and then transported to a rectangular sample pool;

[0022] A fly ash sample having a standard weight and uniform thickness is generated in the rectangular sample pool through a sample valve and a sample gate;

[0023] Continuously collecting microscopic images of the fly ash sample passing through a microscope, receiving light transmittance data corresponding to the fly ash sample passing through an infrared receiver, and recording the ratio of the fly ash weight blown into the fly ash collection pipe by the fixed frequency fan to the fly ash weight entering the fly ash collection bag;

[0024] The collected microscopic image time series, infrared transmittance time series, fly ash weight ratio, boiler operation data and actual carbon content of the fly ash sample are used as training data to train a preset deep learning network model;

[0025] The microscopic images and fly ash physical property parameters collected in real time during the current time period are obtained, and the real-time collected data are input into the trained deep learning network model to predict the carbon content of the fly ash corresponding to the current time period.

[0026] In addition, the online soft measurement method for carbon content in fly ash based on multi-scale coupling in the embodiment of the present application also has the following additional technical features:

[0027] Optionally, in some embodiments, the boiler operation data includes: operating parameters of the boiler and coal quality parameters of the coal burned in the boiler; the deep learning network model includes: an input layer, multiple hidden layers and an output layer.

[0028] Optionally, in some embodiments, measuring the actual carbon content of the fly ash sample includes: measuring the actual carbon content of the fly ash sample by a combustion method; training the preset deep learning network model includes: taking the microscopic image time series, the infrared transmittance time series, the fly ash weight ratio, and the boiler operation data as input data, and taking the actual carbon content of the fly ash sample as output data, and training the preset deep learning network model through a deep learning algorithm until the error value of the model is less than the expected value.

[0029] To achieve the above objectives, the third aspect of the present invention proposes a non-temporary computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, it implements the online soft measurement method of fly ash carbon content based on multi-scale coupling as described in any of the second aspect embodiments above.

[0030] The technical solution provided by the embodiments of the present application brings at least the following beneficial effects:

[0031] The present application collects microscopic image time series through a microscope and a light source, and also collects infrared transmittance time series and physical property characteristics such as weight ratio through other physical property measurement methods, and then measures the carbon content of fly ash by combining continuous microscopic image characteristics and physical property characteristics, and uses the trained neural network model to dynamically analyze the measurement data to predict the carbon content of fly ash. Thus, the present application realizes soft measurement of fly ash carbon content based on dynamic analysis at a microscopic perspective, takes into account the microscopic image of fly ash at a time scale, can avoid the influence of uneven distribution of fly ash carbon on measurement accuracy in a single image, solves the problem of different patterns presented in microscopic images due to uneven sample thickness, and ensures the accuracy of measurement. Therefore, the present application can quickly measure the carbon content of fly ash on the basis of ensuring measurement accuracy, and improves the speed and accuracy of fly ash carbon content measurement.

[0032] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0034] Figure 1A schematic diagram of the structure of an online soft measurement device for carbon content in fly ash based on multi-scale coupling proposed in an embodiment of the present application;

[0035] Figure 2 A flowchart of an online soft measurement method for carbon content in fly ash based on multi-scale coupling proposed in an embodiment of the present application;

[0036] Figure 3 A schematic diagram of a fly ash carbon content prediction principle proposed in an embodiment of the present application;

[0037] Figure 4 A schematic diagram of the structure of a deep neural network prediction model proposed in an embodiment of the present application. DETAILED DESCRIPTION

[0038] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0039] It should be noted that the methods or devices for directly and quickly measuring the carbon content of fly ash in the related art often have a series of problems in practical applications. For example, for the microwave method, it is greatly affected by the fly ash composition, has poor adaptability to coal types, and needs to be calibrated before measurement. The sampling device is easily blocked by ash, which may cause the detection instrument to be unusable in severe cases. For the optical reflection method, due to the uneven particle size of the fly ash sample during the measurement process, the carbon content of the fly ash sample and the reflection signal show a nonlinear relationship. Therefore, the changes in the ash composition and coal type will greatly interfere with the test results and affect the accuracy of the measurement. For the infrared method, this method may be affected by multiple factors such as ambient temperature, humidity, fly ash particle size and chemical composition, resulting in unstable measurement accuracy. For the γ radiation method, it uses radioactive materials as radiation sources, which has radiation safety issues. At the same time, due to the absorption and scattering of γ rays by fly ash, the measurement accuracy may be affected. For the photoacoustic effect method, it may be affected by multiple factors such as light source intensity, optical properties of fly ash, ambient temperature and air pressure, resulting in unstable measurement accuracy. Based on the above problems, various measurement methods in related technologies cannot meet the needs of practical applications.

[0040] In order to measure the carbon content of fly ash quickly and accurately in practical applications, some relevant embodiments begin to use soft measurement to measure fly ash. Soft measurement uses the differences in visual and physical properties of fly ash at different carbon contents to infer the actual carbon content.

[0041] However, the soft measurement method of the carbon content of fly ash in the relevant embodiments is generally measured under macroscopic conditions. Since the characteristics of fly ash are extremely unclear under macroscopic conditions and the benefits are affected by the environment, the measurement accuracy of the soft measurement method of the carbon content of fly ash in the relevant embodiments still cannot meet the requirements.

[0042] To this end, the present application proposes an online soft measurement device and method for fly ash carbon content based on multi-scale coupling, which measures the carbon content of fly ash by combining continuous microscopic image features with the physical properties of fly ash, thereby improving the accuracy and speed of fly ash carbon content measurement.

[0043] The following describes an online soft measurement device and method for carbon content in fly ash based on multi-scale coupling according to an embodiment of the present application with reference to the accompanying drawings.

[0044] Figure 1 The schematic diagram of the structure of an online soft measurement device for carbon content in fly ash based on multi-scale coupling proposed in the embodiment of the present application is as follows: Figure 1 As shown, the device includes: a sampling tube 1, a flue 2, a rectangular sample pool 7, a sample valve 8, a sample gate 9, a transparent conveyor belt 11, a microscope 12, a light source 13, an infrared receiver 14, a fixed frequency fan 15, a fly ash collection tube 16, a fly ash collection bag 18, an infrared transmitter 19, a background baffle 20 and a computing module 21.

[0045] Specifically, the sampling tube 1 is installed in the downstream area of ​​the flue 2, the sampling tube 1 is used to sample the fly ash in the flue 2, and the rectangular sample pool 7 is used to store the collected fly ash.

[0046] As an example, the sampling tube 1 is placed in a stable area downstream of the flue 2 to sample the fly ash flowing in the flue 2. The diameter of the sampling tube 1 may be 20 mm, and its rear end is connected to a device for processing and storing the collected fly ash.

[0047] In one embodiment of the present application, Figure 1 As shown, the device also includes: a sampling valve 3, an exhaust pipe 4, a cyclone separator 5 and a separator outlet 6. The output end of the sampling tube 1 is connected to the input end of the sampling valve 3, and the output end of the sampling valve 3 is connected to the input end of the cyclone separator 5. The cyclone separator 5 is used to separate the gas and solid in the collected fly ash. The separator outlet 6 is connected between the cyclone separator 5 and the rectangular sample pool 7. The separated solid fly ash enters the rectangular sample pool 7 through the separator outlet 6 under the action of centrifugal force.

[0048] Specifically, in this embodiment, the rear end of the sampling tube 1 is also connected to a cyclone separator 5 through a sampling valve 3, and the cyclone separator 5 is used to concentrate and separate the fly ash collected by the sampling tube 1. When concentrating, the cyclone separator 5 separates the gas and solid phases in the sample, so that the solid phase enters the rectangular sample pool 7 from the separator outlet 6 under the action of centrifugal force and is stored, while the gas can be discharged from the exhaust pipe 4. The purpose of concentrating the fly ash is achieved by screening out the gas in the fly ash.

[0049] The sample valve 8 is equipped with a gravity sensor. When the weight of the fly ash falling from the rectangular sample pool 7 reaches a preset weight, the sample valve 8 opens to transport the fly ash to the transparent conveyor belt 11 below; the transparent conveyor belt 11 is used to drive the fly ash to move, and the sample gate 9 is used to cut the passing fly ash into fly ash samples with uniform thickness.

[0050] As an example, the rectangular sample pool 7 is a rectangular hollow pool with an opening on the top, and the outlet of the separator outlet 6 can extend into the interior of the rectangular sample pool 7. A sample valve 8 that can be opened is provided in the middle of the sample pool. The sample valve 8 is arranged on the side wall of the rectangular sample pool 7 and is normally closed by default. The sample valve 8 includes two mechanical structures, a baffle and a valve, and a weight sensor is also provided on the sample valve 8. The weight sensor can detect the weight of the fly ash falling from the separator outlet 6 onto the baffle, and trigger the sample valve 8 to open when the falling material reaches a preset weight, thereby accurately sending the fly ash to the transparent conveyor belt 11 below.

[0051] For example, in a specific implementation, the rectangular sample pool 7 can be set to have a width of 60 mm, a length of 100 mm and a height of 60 mm, and the baffle of the sample valve 8 is set at a height of 30 mm, and a weight sensor is provided above it. When the weight of the falling material at the separator outlet 6 is greater than 200 g, the valve opens to allow the fly ash to fall onto the transparent conveyor belt 11 below under gravity.

[0052] exist Figure 1 In the example shown, the transparent conveyor belt 11 drives the falling fly ash to move to the right, and is cut to form a fly ash sample with uniform thickness when passing through the sample gate 9. For example, the height of the sample gate 9 can be set to 2 mm, so that the fly ash becomes a fly ash sample with a thickness of 2 mm after passing through the sample gate 9, so as to facilitate subsequent testing.

[0053] It is understandable that the image at the microscopic scale is also related to the thickness of the fly ash sample. When the thickness is uneven, the fly ash sample presents different patterns under the microscopic image. In order to avoid the influence of thickness on the measurement, the present application sets a weight sensor and a sample gate 9 to ensure that the weight and thickness of the fly ash sample measured subsequently are consistent each time, and the overall thickness of the generated single fly ash sample is uniform.

[0054] In one embodiment of the present application, the device also includes: a cleaning baffle, which is located below the rectangular sample pool 7, and the bottom of the cleaning baffle is made of brush material; the cleaning baffle is used to block the background and clean impurities on the conveyor belt to be introduced into the rectangular sample pool 7.

[0055] Specifically, in this embodiment, a dust cleaning baffle ( Figure 1 (not shown in the figure), the specific installation position of the cleaning baffle can be adjusted according to the actual situation. For example, it can be set on the lower left side of the rectangular sample pool 7. The bottom of the cleaning baffle is a brush material. By setting the brush material, it can not only block the background, but also effectively clean the impurities on the conveyor belt that is about to enter the rectangular sample pool 7, so as to determine the purity of the fly ash sample to be measured later.

[0056] It should be noted that when the transparent conveyor belt 11 is running, it enters a part of the rectangular sample pool 7 to carry the fallen fly ash. In actual applications, there may be other impurities on the surface of the transparent conveyor belt 11. For this reason, the brush material at the bottom of the cleaning baffle in this embodiment can clean the impurities on the surface of the transparent conveyor belt 11 to ensure that only the fallen fly ash is contained on the transparent conveyor belt 11. In addition, since the microscope 12 is close to the rectangular sample pool 7, and the microscope 12 needs to avoid interference from factors such as light from other directions and the background under the transparent conveyor belt 11 when measuring, the brush material at the bottom of the cleaning baffle can also play a role in shielding the background.

[0057] In one embodiment of the present application, the device further includes: a dust cleaner 10, which is located below the rectangular sample pool 7 and is used to clean the conveyor belt to be introduced into the rectangular sample pool 7 in combination with the dust cleaning baffle; and a conveyor belt roller 17, which is used to drive the transparent conveyor belt 11 to run. Specifically, in this embodiment, the dust cleaner 10 can clean the transparent conveyor belt 11 area to be introduced into the rectangular sample pool 7 in combination with the dust cleaning baffle in the above embodiment through its dust cleaning mechanical structure, so as to further improve the dust cleaning effect.

[0058] The microscope 12 is used to continuously collect microscopic images of the fly ash sample with the assistance of the light source 13 and the background baffle 20; the infrared receiver 14 is located behind the microscope 12, and the infrared transmitter 19 and the infrared receiver 14 are used to continuously collect the light transmittance data of the fly ash sample; the fly ash collection tube 16 and the fly ash collection bag 18 are used to record the fly ash weight ratio of the two outlets.

[0059] Specifically, under the action of the transparent conveyor belt 11, the fly ash samples prepared by the sample gate 9 are successively passed through the microscope 12 and the infrared transmitter 19 and other equipment arranged behind the rectangular sample pool 7 for microscopic image observation and various fly ash physical property analyses, including collecting microscopic images through the microscope 12, performing infrared light inspection through the infrared transmitter 19 and the infrared receiver 14, and blowing the fly ash to the fly ash collection pipe 16 and the fly ash collection bag 18 under the action of the fan to detect the fly ash viscosity.

[0060] As an example, a microscope 12 is installed behind the rectangular sample pool 7, and its lens is aimed at the transparent conveyor belt 11 that drives the fly ash sample to pass through, so as to capture a clear microscopic image. A background baffle 20 is specially set below the conveyor belt at the bottom of the microscope 12 to optimize the background effect of the image. At the same time, a dedicated light source 13 is installed on the side of the microscope 12 to illuminate the fly ash sample currently being observed to ensure the clarity of image acquisition.

[0061] In the specific implementation, in this example, the lens of the microscope 12 is set at a position 5 mm away from the transparent conveyor belt, and the lens adopts a 200-fold lens. A 60W light source is set on the side of the lens to illuminate the sample.

[0062] In one embodiment of the present application, the microscope 12 further includes: a data transmission module, which is used to transmit the collected microscopic image in real time to the storage device of the computing module 21. Specifically, in this embodiment, the microscope 12 is also equipped with a data transmission module, which can transmit the currently captured microscopic image data in real time to an image storage digital device, such as the memory of the computing module 21, through wireless transmission, so as to facilitate subsequent analysis and processing.

[0063] The infrared transmitter 19 is installed below the transparent conveyor belt 11 and behind the microscope 12. At the same time, an infrared receiver 14 is arranged at a corresponding position above the infrared transmitter 19. The infrared receiver 14 is used to receive and record the light transmission data after the infrared light irradiates the fly ash sample. The light transmission data can reflect the reflectivity of the fly ash sample.

[0064] In specific implementation, as an example, the infrared transmitter 19 is installed 10 mm below the transparent conveyor belt, 120 mm behind the microscope 12, and an infrared receiver 14 is provided 10 mm above the vertical conveyor belt to receive infrared light transmittance data. In this example, the infrared receiver 14 is also used to transmit the collected light transmittance data to the storage device of the computing module 21 through wireless transmission, so as to facilitate subsequent analysis and processing.

[0065] Continue to refer to Figure 1In the example shown, the fly ash sample that has been subjected to infrared detection is blown into a fly ash collection pipe 16 and a fly ash collection bag 18 by a fixed-frequency blower 15 .

[0066] As an example, the fixed-frequency fan 15 is installed behind the infrared receiver 14, and its air outlet is aligned above the transparent conveyor belt 11. When the fly ash sample passes by, the fan will blow it into the fly ash collection pipe 16 at a fixed frequency to ensure the smooth collection of the sample. The fly ash sample that is not blown into the fly ash collection pipe 16 continues to enter the fly ash collection bag 18 driven by the transparent conveyor belt 11. A weight sensor and a data transmission device are set in the fly ash collection pipe 16 and the fly ash collection bag 18. The weight of the fly ash currently falling in is detected by the weight sensor, and the currently detected weight data is transmitted to the memory of the operation module 21 through the data transmission device, so as to facilitate subsequent analysis and processing, and calculate the weight ratio of the fly ash currently falling into the two outlets of the fly ash collection pipe 16 and the fly ash collection bag 18.

[0067] The operation module 21 is connected to each acquisition device for communication. The operation module 21 is used to dynamically analyze the continuously acquired microscopic images and fly ash physical property parameters to output the carbon content of the fly ash.

[0068] Specifically, each collection device includes the microscope 12, the infrared receiver 14, the fly ash collection tube 16 and the fly ash collection bag 18, etc. Through the data transmission method in the above embodiment, the operation module 21 is connected to each collection device in communication to receive the data collected by each collection device ( Figure 1 In the figure, the fly ash sample is connected to the microscope 12 and the infrared receiver 14 for exemplary description. Then, the operation module 21 analyzes the data collected in real time to predict the carbon content of the current fly ash sample.

[0069] It should be noted that, since fly ash is a product of condensation in the furnace, its distribution at the microscopic scale is extremely uneven, and the image characteristics of the same ash sample at 200 times may be completely different. Therefore, the measurement method based on only one static image for analysis will result in poor accuracy. To this end, the present application can continuously collect microscopic images and fly ash physical parameters of each fly ash sample generated at different times over a period of time according to the collection method of each of the above-mentioned devices, obtain the time series data of each test result in the time period, and then the calculation module 21 performs dynamic analysis based on the time series data to predict the carbon content of the fly ash sample in the time period.

[0070] In one embodiment of the present application, the operation module 21 is equipped with a trained deep learning network model, which is used to predict the fly ash carbon content corresponding to the current time period based on the received microscopic image time series, infrared transmittance time series, fly ash weight ratio and boiler operation data collected in real time during the current time period.

[0071] Therefore, the device combines microscopic image observation and fly ash physical property analysis to achieve online measurement of fly ash carbon content. Measurement through this device not only ensures the measurement accuracy, but also significantly improves the measurement speed.

[0072] In summary, the online soft measurement device for carbon content in fly ash based on multi-scale coupling implemented in the present application collects microscopic image time series through a microscope and a light source, and also collects physical properties such as infrared transmittance time series and weight ratio through other physical property measurement methods, and then measures the carbon content in fly ash by combining continuous microscopic image features with physical property features. Thus, the device realizes soft measurement of carbon content in fly ash based on dynamic analysis at a microscopic perspective, takes into account the microscopic image of fly ash at a time scale, and can avoid the influence of uneven carbon distribution of fly ash on measurement accuracy in a single image, solves the problem of different patterns presented in a microscopic image due to uneven sample thickness, and ensures the accuracy of measurement. As a result, the device can quickly measure the carbon content in fly ash on the basis of ensuring measurement accuracy, thereby improving the speed and accuracy of measuring the carbon content in fly ash.

[0073] In order to more clearly illustrate the complete testing process of the online soft measurement device for the carbon content in fly ash of the present application, as well as the specific implementation process of analyzing the measurement data to obtain the measurement results, an online soft measurement method for the carbon content in fly ash based on multi-scale coupling proposed in the embodiment of the present application is described in detail below. This method is applied to the online soft measurement device for the carbon content in fly ash based on multi-scale coupling in the above embodiment, that is, the measurement method of this embodiment is realized by performing relevant control on the measurement device in the above embodiment. The various measurement devices involved in this method are as described in the above embodiment and will not be repeated here. The execution subject of the online soft measurement of the carbon content in fly ash based on multi-scale coupling in the embodiment of the present application can be the various modules in the above device embodiment, such as the operation module.

[0074] Figure 2 A flowchart of an online soft measurement method for carbon content in fly ash based on multi-scale coupling proposed in an embodiment of the present application is shown in FIG. Figure 2 As shown, the method comprises the following steps:

[0075] Step S101: fly ash is collected in the flue through a sampling tube, and the collected fly ash is concentrated by a cyclone separator and then transported to a rectangular sample pool.

[0076] Step S102: Generate a fly ash sample with a standard weight and uniform thickness in a rectangular sample pool through a sample valve and a sample gate.

[0077] Specifically, a certain amount of fly ash is collected in the flue through a sampling tube, and is concentrated by a cyclone separator and then sent to a rectangular sample pool. The weight of the fly ash in the rectangular sample pool is monitored to ensure that the amount of fly ash sample generated each time is consistent.

[0078] Furthermore, when the fly ash on the sample valve reaches a preset weight, the sample valve is opened to allow the sample to fall onto the transparent conveyor belt below. As the transparent conveyor belt runs, the sample is carried out of the sample valve port, thereby making a fly ash sample with uniform thickness. Since the weight of the fly ash falling each time the sample valve is opened is the same, the weight of each fly ash sample generated during the collection period is a standard value.

[0079] Step S103: continuously collect microscopic images of the passing fly ash samples through a microscope, receive the light transmittance data corresponding to the passing fly ash samples through an infrared receiver, and record the ratio of the fly ash weight blown into the fly ash collection pipe by the fixed-frequency fan to the fly ash weight entering the fly ash collection bag.

[0080] Specifically, the generated fly ash samples are driven by a transparent conveyor belt and pass through a microscope and an infrared receiver in turn to reach two outlets. When passing through the microscope, the microscope captures the microscopic image of the fly ash passing through. Then, the infrared receiver receives the infrared transmittance of the fly ash passing through. Then, the fly ash samples passing through are blown into the outlet of the fly ash collection pipe through a fixed-frequency fan, and the remaining fly ash that has not entered the pipe enters the outlet of the fly ash collection bag below, and the fly ash weights of the two outlets are recorded.

[0081] Step S104: The collected microscopic image time series, infrared transmittance time series, fly ash weight ratio, boiler operation data and actual carbon content of the fly ash sample are used as training data to train the preset deep learning network model.

[0082] Specifically, after the fly ash samples are tested as above, continuous microscopic image data, infrared reflection data and sample weight ratio data within the sampling period can be obtained, that is, the microscopic image time series and infrared transmittance time series are obtained, among which the weight ratio data of the two outlets can reflect the viscosity of the fly ash samples. Combined with the boiler operation data obtained within the current sampling period, it is used as the input data in the deep learning training process.

[0083] Furthermore, the actual carbon content of the fly ash sample is obtained through actual detection and calibration, and the actual carbon content of the fly ash is used as the output data in the deep learning training process.

[0084] Furthermore, by repeatedly executing steps S101 to S104, measurement data within multiple sampling periods can be obtained, and the process is repeated until enough data required for model training is accumulated. Then, the input and output data obtained are used as training data to train the preset deep learning network model.

[0085] The type of the preset deep learning network model can be determined according to the actual prediction requirements. For example, a LightGBM model or an XGBoost model can be used. The principle of this application to obtain various test data for measuring the carbon content of fly ash is as follows: Figure 3 As shown, based on this process, a deep learning network model is trained to predict the carbon content in fly ash.

[0086] In one embodiment of the present application, the boiler operation data is plant-level operation data that affects the carbon content of fly ash, including boiler operation parameters and coal quality parameters of coal burned in the boiler, etc. When obtaining the actual carbon content of the fly ash sample, the actual carbon content of the fly ash sample can be measured by the combustion method in the relevant embodiment. For example, the actual carbon content of the fly ash sample can be detected by using the loss on ignition test in the laboratory analysis in the prior art, and the detection result is used as the true value.

[0087] In one embodiment of the present application, the deep learning network model includes: an input layer, multiple hidden layers and an output layer, and the preset deep learning network model is trained, including: taking the microscopic image time series, infrared transmittance time series, fly ash weight ratio, and boiler operation data as input data, and taking the actual carbon content of the fly ash sample as output data, and training the preset deep learning network model through a deep learning algorithm until the error value of the model is less than the expected value.

[0088] Specifically, if Figure 4 As shown, the various types of input data obtained are input into the input layer of the network model, and the output data is input into the output layer of the network model. The network model is trained through the relevant deep learning training algorithm, and the parameters of the hidden layer are optimized until the training error of the ash carbon content is less than the expected value. The training is stopped and the model training is completed. Among them, the error value of the model is the difference between the predicted value output by the model and the actual carbon content corresponding to the input data. The expected value is determined according to actual factors such as the prediction accuracy requirements. The specific model training process can refer to the training process of the deep learning network model in the relevant technology, and this application does not limit this.

[0089] It should be noted that the process from step S101 to step S104 can be performed in advance in an offline stage, and the trained deep learning network model with the function of estimating the carbon content of fly ash can be used for subsequent actual measurement applications.

[0090] Step S105: Acquire the microscopic images and fly ash physical property parameters collected in real time during the current period, input the real-time collected data into the trained deep learning network model, and predict the carbon content of the fly ash corresponding to the current period.

[0091] Specifically, the trained model is put into use in the production process. When the online soft measurement of the fly ash carbon content is actually carried out, the fly ash to be measured in the current time period is first detected according to the fly ash collection and detection process of steps S101 to S103 above, and the real-time collected microscopic image time series, infrared transmittance time series, sample weight ratio data and boiler operation data are obtained. The real-time collected data is then input into the trained deep learning network model to predict the fly ash carbon content corresponding to the current time period.

[0092] For example, in the boiler production process, the trained deep learning network model is used for inference to complete the rapid detection of fly ash carbon content. In actual use, after a fixed interval, a 10-minute timed sampling is carried out. When the sample is taken out from the flue and enters the fly ash collection pipe, the measurement result of the fly ash carbon content can be quickly obtained within 10 seconds.

[0093] In summary, the online soft measurement method of fly ash carbon content based on multi-scale coupling in the embodiment of the present application measures the fly ash carbon content by combining continuous microscopic image features and physical property features, thereby realizing soft measurement of fly ash carbon content based on dynamic analysis at a microscopic perspective, taking into account the fly ash microscopic image at a time scale, and being able to avoid the influence of uneven fly ash carbon distribution on measurement accuracy under a single image, solving the problem of different patterns presented under a microscopic image due to uneven sample thickness, and ensuring the accuracy of the measurement. Therefore, the method can quickly measure the fly ash carbon content on the basis of ensuring measurement accuracy, thereby improving the speed and accuracy of the fly ash carbon content measurement.

[0094] In order to implement the above-mentioned embodiment, the present application also proposes a non-temporary computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the online soft measurement method of fly ash carbon content based on multi-scale coupling as proposed in the aforementioned second aspect embodiment of the present application.

[0095] It should be noted that it should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0096] In addition, in the description of the present application, the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc., indicating the orientation or position relationship are based on the orientation or position relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0097] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0098] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements, unless otherwise clearly defined. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0099] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0100] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in the field can change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. An online soft measurement device for carbon content in fly ash based on multi-scale coupling, characterized in that: include: Sampling tube, flue, rectangular sample pool, sample valve, sample gate, transparent conveyor belt, microscope, light source, infrared receiver, fixed frequency fan, fly ash collection tube, fly ash collection bag, infrared transmitter, background baffle and computing module; among them, The sampling tube is installed in the downstream area of ​​the flue, the sampling tube is used to sample the fly ash in the flue, and the rectangular sample pool is used to store the collected fly ash; The sample valve is equipped with a gravity sensor. When the weight of the fly ash falling from the rectangular sample pool reaches a preset weight, the sample valve opens to transport the fly ash to the transparent conveyor belt below; the transparent conveyor belt is used to drive the fly ash to move, and the sample gate is used to cut the fly ash passing through into fly ash samples with uniform thickness; The microscope is used to continuously collect microscopic images of the fly ash sample with the assistance of the light source and the background baffle; the infrared receiver is located behind the microscope, and the infrared transmitter and the infrared receiver are used to continuously collect light transmittance data of the fly ash sample; the fly ash collection pipe and the fly ash collection bag are used to record the fly ash weight ratio of the two outlets; The operation module is in communication connection with each acquisition device, and is used for dynamically analyzing the continuously acquired microscopic images and fly ash physical property parameters to output the carbon content of the fly ash.

2. The device according to claim 1, characterized in that It also includes: a sampling valve, an exhaust pipe, a cyclone separator and a separator outlet; wherein, The output end of the sampling tube is connected to the input end of the sampling valve, and the output end of the sampling valve is connected to the input end of the cyclone separator, and the cyclone separator is used to separate the gas and solid in the collected fly ash; The separator outlet is connected between the cyclone separator and the rectangular sample pool, and the separated solid fly ash enters the rectangular sample pool through the separator outlet under the action of centrifugal force.

3. The device according to claim 1, characterized in that Also includes: A dust cleaning baffle, the dust cleaning baffle is located below the rectangular sample pool, and the bottom of the dust cleaning baffle is made of brush material; The dust cleaning baffle is used to shield the background and clean the impurities on the conveyor belt to be fed into the rectangular sample pool.

4. The device according to claim 3, characterized in that Also includes: An ash cleaner, the ash cleaner is located below the rectangular sample pool, and the ash cleaner is used to clean the conveyor belt to be introduced into the rectangular sample pool in combination with the ash cleaning baffle; The conveyor belt roller is used to drive the transparent conveyor belt to run.

5. The device according to claim 1, characterized in that The microscope comprises: A data transmission module, the data transmission module is used to transmit the collected microscopic image to the storage device of the computing module in real time; The infrared receiver is also used to transmit the collected light-transmitting data to the storage device of the computing module.

6. The device according to claim 1, characterized in that The computing module is equipped with a trained deep learning network model, which is used to predict the fly ash carbon content corresponding to the current time period based on the real-time collected microscopic image time series, infrared transmittance time series, the fly ash weight ratio and boiler operation data.

7. An online soft measurement method for fly ash carbon content based on multi-scale coupling, characterized in that: Applied to the online soft measurement device for carbon content in fly ash based on multi-scale coupling as claimed in any one of claims 1 to 6, the measurement method comprises the following steps: Fly ash is collected in the flue through a sampling tube, and the collected fly ash is concentrated by a cyclone separator and then transported to a rectangular sample pool; A fly ash sample having a standard weight and uniform thickness is generated in the rectangular sample pool through a sample valve and a sample gate; Continuously collecting microscopic images of the fly ash sample passing through a microscope, receiving light transmittance data corresponding to the fly ash sample passing through an infrared receiver, and recording the ratio of the fly ash weight blown into the fly ash collection pipe by the fixed frequency fan to the fly ash weight entering the fly ash collection bag; The collected microscopic image time series, infrared transmittance time series, fly ash weight ratio, boiler operation data and actual carbon content of the fly ash sample are used as training data to train a preset deep learning network model; The microscopic images and fly ash physical property parameters collected in real time during the current time period are obtained, and the real-time collected data are input into the trained deep learning network model to predict the carbon content of the fly ash corresponding to the current time period.

8. The method according to claim 7, characterized in that The boiler operation data includes: boiler operation parameters and coal quality parameters of the coal burned in the boiler; The deep learning network model includes: an input layer, multiple hidden layers and an output layer.

9. The method according to claim 8, characterized in that Measuring the actual carbon content of the fly ash sample includes: Measuring the actual carbon content of the fly ash sample by an ignition method; The training of the preset deep learning network model includes: The microscopic image time series, the infrared transmittance time series, the fly ash weight ratio, and the boiler operation data are used as input data, and the actual carbon content of the fly ash sample is used as output data. The preset deep learning network model is trained through a deep learning algorithm until the error value of the model is less than the expected value.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, an online soft measurement method for carbon content in fly ash based on multi-scale coupling is implemented as described in any one of claims 7 to 9.

Images