Boiler hearth cleaning control system based on CCD imaging recognition

Abstract

The invention belongs to the technical field of boiler furnace cleaning control, and particularly discloses a boiler furnace cleaning control system based on CCD imaging recognition, which comprises an image acquisition module, a pollution recognition module, a grade confirmation module, a cleaning control module, a residue analysis module and a feedback control terminal, according to the invention, through surface texture features such as thickness, color and texture in the visual image, contour type determination and area calculation of ash deposition and coking are realized, the problem of lack of visual monitoring at present is solved, key parameters such as ash deposition coverage rate and coking thickness are acquired in real time, quantitative data support is provided for subsequent cleaning control, and the method is suitable for popularization and application. And meanwhile, the pollution level is dynamically upgraded by combining the temperature characteristic of the infrared thermodynamic diagram based on the analysis of the coking thickness and the ash deposition coverage rate, so that the cleaning efficiency is improved, and the risk of equipment overtemperature caused by the misjudgment of the pollution level is avoided through the fusion analysis of the infrared temperature data and the visual characteristic.

Description

Technical Field

[0001] The invention belongs to the technical field of boiler furnace cleaning control and relates to a boiler furnace cleaning control system based on CCD imaging recognition. Background Art

[0002] The boiler furnace serves as the core combustion space of coal-fired, oil-fired, or gas-fired boilers. Its inner walls, including water-cooled walls, transfer heat released by fuel combustion to the working fluid through heat exchange. During boiler operation, ash, incompletely burned particles, and other substances in the fuel gradually deposit on the heating surfaces, forming ash deposits or coke. Ash deposits reduce the heat transfer efficiency of the heating surfaces, resulting in decreased boiler thermal efficiency and increased energy consumption. Coke can cause excessive heating surface temperature and tube bursts due to excessive thickness, or cause abnormal furnace pressure due to ash accumulation clogging the flue, seriously threatening the safe and stable operation of the boiler. Therefore, precise cleaning control of the boiler furnace is key to ensuring efficient and safe operation of the equipment.

[0003] For example, Chinese invention patent publication number CN108716690A discloses a clean industrial boiler system whose core structure includes a rectifier unit, an inverter unit, an induced draft fan, a blower, and a PLC control unit. This system achieves furnace control through the following methods: Pressure control logic: The PLC collects furnace pressure signals in real time and uses intelligent algorithms to adjust the operating frequency of the induced draft fan and blower, controlling the air intake and exhaust volumes to stabilize the furnace negative pressure. Hardware configuration: A rectifier unit and dual inverter units with a common DC bus are used to drive the induced draft fan and blower, and communication is achieved via a DP bus. Mechanical structure: The adjustable windshield is fixed at its maximum angle to simplify the mechanical adjustment structure.

[0004] The above existing technologies have the following deficiencies: 1. Currently, the furnace pressure is only controlled by PLC, and visual monitoring of the pollution status inside the furnace is not involved. Key parameters such as the ash thickness and coking area of ​​the furnace heating surface cannot be obtained in real time, making it difficult to judge the type and severity of pollution. As a result, the cleaning operation lacks data support, and over-cleaning or under-cleaning may occur.

[0005] 2. Currently, the pressure is controlled only by fixing the wind deflector angle and adjusting the fan frequency. No differentiated cleaning strategy is formulated according to the pollution status. The cleaning operation relies on preset programs and cannot dynamically adjust the purge pressure, duration and other parameters according to the actual pollution level. The cleaning efficiency is low and energy consumption is seriously wasted. Summary of the Invention

[0006] In view of this, in order to solve the problems raised in the above background technology, a boiler furnace cleaning control system based on CCD imaging recognition is proposed.

[0007] The objectives of the present invention can be achieved through the following technical solutions: The present invention provides a boiler furnace cleaning control system based on CCD imaging recognition, including: an image acquisition module for dividing the interior of the boiler furnace into sub-areas according to the furnace structure characteristics, and collecting visual images and infrared thermal maps of each sub-area through CCD.

[0008] The pollution identification module is used to extract surface texture features from the visual image and divide the sub-regions into regional types.

[0009] The level confirmation module is used to analyze the coking thickness and ash coverage based on the area type of each sub-area, and determine the pollution level of each sub-area in combination with the infrared thermal map.

[0010] The cleaning control module is used to match a cleaning method based on the pollution level and perform cleaning control according to the cleaning method.

[0011] The residual analysis module is used to trigger the secondary image acquisition of each sub-area after cleaning is completed, obtain the secondary visual image of each sub-area, and compare the features with the reference cleaning state image to obtain each residual area.

[0012] The feedback control terminal is used to determine the residual pollution level based on the residual coke thickness and residual ash coverage of each residual area, and then trigger secondary cleaning in combination with the position coordinates of each residual area and provide corresponding feedback.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention uses surface texture features such as thickness, color, and texture in visual images to determine the contour type and calculate the area of ​​dust and coke deposits, solving the current problem of lack of visual monitoring. Key parameters such as dust coverage and coke thickness are obtained in real time, providing quantitative data support for subsequent cleaning control.

[0014] (2) The present invention improves the accuracy of contamination identification and avoids the problem of over-cleaning or under-cleaning through secondary verification of contour type and mixed contour area allocation algorithm.

[0015] (3) The present invention improves the cleaning efficiency by dynamically upgrading the pollution level based on the analysis of coking thickness and dust coverage combined with the temperature characteristics of infrared thermal images. At the same time, the fusion analysis of infrared temperature data and visual characteristics reduces the misjudgment rate of high-temperature coking and avoids the risk of equipment overheating due to misjudgment of pollution levels.

[0016] (4) The present invention locates the residual area by comparing the secondary visual image with the baseline cleaning state image, and triggers secondary cleaning according to the residual parameters, thereby solving the problem of no verification of the current cleaning effect. The cleaning compliance rate is improved through the quantitative evaluation of the residual coke thickness and the ash coverage rate.

[0017] (5) The present invention calculates the coking and ash accumulation levels of the mixed area separately to obtain the comprehensive pollution level, and corrects the level by combining the infrared temperature and low-temperature area, which effectively solves the current problem of lack of mixed pollution treatment and improves the cleaning efficiency of the mixed pollution area. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0019] Figure 1 This is a schematic diagram of the connection of various modules of the system of the present invention.

[0020] Figure 2 This is a connection diagram of the regional type division steps of the present invention.

[0021] Figure 3 Schematic diagram of the connection of the residual area analysis step of the present invention. DETAILED DESCRIPTION

[0022] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0023] See also Figure 1 As shown, the present invention provides a boiler furnace cleaning control system based on CCD imaging recognition, which includes: an image acquisition module, a pollution recognition module, a level confirmation module, a cleaning control module, a residue analysis module and a feedback control terminal.

[0024] In the above, the pollution identification module is connected to the image acquisition module and the level confirmation module respectively, the cleaning control module is connected to the level confirmation module and the residue analysis module respectively, and the residue analysis module is also connected to the feedback control terminal.

[0025] The image acquisition module is used to divide the interior of the boiler furnace into sub-areas according to the furnace structure characteristics, and to acquire visual images and infrared thermal images of each sub-area through CCD.

[0026] Exemplarily, the dividing of the sub-regions into regional types includes dividing the furnace into three layers of regions, namely, upper, middle and lower regions, along the vertical direction of the furnace based on the spatial distribution of each structural component in the furnace structural characteristics.

[0027] Each layer area is divided into several sector-shaped areas according to a preset circumferential angle and used as sub-areas.

[0028] The pollution identification module is used to extract surface texture features from the visual image and divide each sub-region into a region type.

[0029] See also Figure 2 As shown, exemplarily, the region type division of each sub-region includes: Q1, determining the contour type of each contour in each sub-region based on the thickness, color and texture of each contour in each sub-region in the surface texture feature.

[0030] Furthermore, the determining of the contour type of each contour in each sub-region includes: Q1-1, extracting a middle value from a thickness range corresponding to the dust accumulation type as a dust accumulation reference thickness.

[0031] It should be added that the method for obtaining the thickness interval corresponding to the dust accumulation type is: extracting the contour thickness of each boiler furnace cleaning from the historical data, and then constructing the thickness interval corresponding to the dust accumulation type

[0032] Q1-2. Calculate the similarity between the thickness of each contour and the dust accumulation reference thickness to obtain the dust accumulation thickness similarity of each contour.

[0033] It should be added that the calculation formula for the dust accumulation thickness similarity is: , where is the similarity of dust accumulation thickness, and are the profile thickness and dust accumulation reference thickness respectively, is the thickness difference of the reference, where The acquisition method is: extract the historical profile thickness difference between the profile thickness of each boiler furnace cleaning and the ash accumulation benchmark thickness from the historical data, and then select the maximum value as the thickness difference for setting the reference.

[0034] Q1-3. The dust color similarity and dust texture similarity of each contour are calculated similarly according to the calculation method of the dust thickness similarity.

[0035] It should be added that the dust color chromaticity value and dust contrast of each contour are extracted from the visual image, and then the dust color similarity and dust texture similarity of each contour are calculated in the same way as the calculation method of the dust thickness similarity.

[0036] Q1-4. Perform weighted fusion calculation on the dust accumulation thickness similarity, dust accumulation color similarity, and dust accumulation texture similarity of each contour to obtain the comprehensive dust accumulation similarity of each contour.

[0037] It should be added that the analysis formula of the dust accumulation comprehensive similarity is: , where is the comprehensive similarity of dust accumulation, and are the dust color similarity and dust texture similarity respectively, 、 and are the weights of dust thickness similarity, dust color similarity and dust texture similarity, respectively. , .

[0038] It should be added that the thickness of dust accumulation is the core indicator for judging the degree of pollution. A thicker dust layer will significantly affect the heat transfer efficiency of the heated surface, and the thickness is directly related to the difficulty of cleaning, so its weight is the largest. The color of the dust accumulation can reflect its composition and deposition time. Color similarity is important for distinguishing dust accumulation from coking, but it is easy to make misjudgments when used as a basis for judgment alone. For example, the oxide layer on the surface of coking may appear gray, so its weight is given to the thickness of the dust accumulation. The texture characteristics of the dust accumulation are greatly affected by the particle size and deposition method. The texture under different working conditions is significantly different, resulting in poor universality, so its weight is the lowest, so it is set. , in order to facilitate analysis, The specific value can be 0.5. The specific value can be 0.3, The specific value can be 0.2.

[0039] Q1-5. The coking comprehensive similarity of each contour is obtained by analyzing the comprehensive similarity of dust accumulation in the same way.

[0040] Q1-6. Compare the comprehensive similarity of dust accumulation and coking of each contour, and select the type corresponding to the high similarity as the initial type of the contour.

[0041] Q1-7. With the center point of the contour as the center of the circle and the preset radius as the radius, the influence range of each contour is circled, and each contour within the influence range is obtained as each influencing contour, and then the contour type is secondary verified based on each influencing contour corresponding to each contour to obtain the contour type of each contour in each sub-area.

[0042] Furthermore, the analysis of each profile type in each sub-area includes: Q1-7-1, based on the initial type of each influencing profile, counting the number of influencing profiles consistent with the profile type and the total number of influencing profiles, and then taking the ratio of the two as the collaborative verification coefficient of the profile.

[0043] Q1-7-2. Compare the collaborative verification coefficient with the preset verification threshold. When the collaborative verification coefficient is greater than or equal to the preset verification threshold, the initial type of the contour is used as the contour type of the contour; otherwise, the mixed type is used as the contour type of the contour, thereby obtaining the contour type of each contour in each sub-area.

[0044] It should be added that the preset verification threshold is a key parameter for judging the reliability of the contour type, and its value range is [0,1]. Its core function is: when the co-verification coefficient of a certain contour is ≥ the threshold, it means that the initial type of the contour has spatial consistency and the judgment result is reliable. If the co-verification coefficient is less than the threshold, it indicates that the contour type is significantly different from the surrounding influencing contours and needs to be judged as a mixed type to avoid misjudgment of a single type. The preset verification threshold is determined by training with historical clean data: a large number of contour samples of known types are collected, their co-verification coefficients are calculated, and the misjudgment rate is set to be less than 5%. The optimal threshold, such as 0.75, is determined by constructing the ROC curve.

[0045] Q2. Extract the area of ​​each contour in each sub-region from the surface texture features, and then combine the contour type to obtain the dust accumulation area and coking area of ​​each sub-region.

[0046] Furthermore, the analysis of the dust accumulation area and coking area of ​​each sub-region includes: Q2-1, extracting each mixed type of profile from the profile type and recording them as each mixed profile.

[0047] Q2-2. Normalize the comprehensive similarity of dust accumulation and the comprehensive similarity of coking of each mixed profile to obtain the dust accumulation ratio and coking ratio of each mixed profile.

[0048] It should be added that the calculation formulas for the normalization processing are: and , where and are the ash accumulation ratio and coking ratio, and They are the comprehensive similarity of dust accumulation and the comprehensive similarity of coking respectively.

[0049] Q2-3. Allocate the area of ​​each mixed profile according to the dust accumulation ratio and the coking ratio to obtain the dust accumulation area and the coking area of ​​each mixed profile.

[0050] Q2-4. According to the method for allocating the dust accumulation area and coking area of ​​each mixed profile, the areas of the profiles with the profile types of dust accumulation and coking are allocated in the same way, thereby obtaining the dust accumulation area and coking area of ​​each profile.

[0051] Q2-5. Sum the dust accumulation area and coking area of ​​each contour in each sub-region to obtain the dust accumulation area and coking area of ​​each sub-region.

[0052] Q2-6. Normalize the dust accumulation area and coking area of ​​each sub-region to obtain the dust accumulation coverage rate and coking coverage rate of each sub-region.

[0053] Q3. Normalize the dust accumulation area and coking area of ​​each sub-region to obtain the dust accumulation coverage rate and coking coverage rate of each sub-region.

[0054] Q4. Determine the area type of each sub-area based on the dust accumulation coverage rate and coking coverage rate of each sub-area.

[0055] Furthermore, the determination of the area type of each sub-area includes: Q4-1, if the dust accumulation coverage of a sub-area is greater than or equal to its preset dust accumulation threshold and the coking coverage is less than its preset coking threshold, then the area type of the sub-area is determined to be dust accumulation.

[0056] Q4-2. If the coking coverage of a sub-region is greater than or equal to its preset coking threshold and the dust accumulation coverage is less than its preset dust accumulation threshold, the region type of the sub-region is determined to be coking.

[0057] Q4-3. If the coking coverage of a sub-region is less than its preset coking threshold and the dust accumulation coverage is less than its preset dust accumulation threshold, the region type of the sub-region is determined to be mixed, and then the region type of each sub-region is obtained.

[0058] The embodiment of the present invention improves the accuracy of contamination identification and avoids the problem of over-cleaning or under-cleaning through secondary verification of contour types and a mixed contour area allocation algorithm.

[0059] It should be noted that the preset ash accumulation threshold is a critical value for determining whether a sub-area is dominated by ash accumulation, and the preset coking threshold is a critical value for determining whether a sub-area is dominated by coking. By quantifying the area ratio of ash accumulation and coking, the continuous pollution state is divided into discrete types, providing a decision-making basis for the cleaning strategy. The threshold is obtained based on historical data. The data on ash accumulation or coking coverage after cleaning under different boiler loads and coal qualities are collected to form a sample set. Based on the sample set, the probability density curve of the ash accumulation coverage rate and the coking coverage rate is drawn, and the inflection point of the curve is taken as the threshold.

[0060] The embodiment of the present invention uses surface texture features such as thickness, color, and texture in visual images to determine the contour type and calculate the area of ​​dust and coke deposits, solving the current problem of lack of visual monitoring. Key parameters such as dust coverage and coke thickness are obtained in real time, providing quantitative data support for subsequent cleaning control.

[0061] The level confirmation module is used to analyze the coking thickness and dust accumulation coverage based on the area type of each sub-area, and determine the pollution level of each sub-area in combination with the infrared thermal map.

[0062] Exemplarily, determining the pollution level of each sub-area includes: if the area type of a sub-area is coking, extracting the maximum coking thickness from the surface texture features as the baseline coking thickness of the sub-area, and matching it with the coking thickness interval corresponding to each coking level to obtain the coking pollution level of the sub-area.

[0063] It should be added that the coke thickness range corresponding to the coke grade is obtained based on the setting of historical data. The coke thickness data of the boiler under different operating conditions are collected, and the probability density curve of the coke thickness is drawn. The inflection point or quantile of the curve is taken as the interval boundary. For example: in the historical data of a boiler, the coke thickness ≤1mm accounts for 60%, 1-3mm accounts for 30%, and >3mm accounts for 10%. The interval is set as: low level is 0-1mm, medium level is 1-3mm, and high level is >3mm.

[0064] The maximum temperature of the sub-area is obtained from the infrared thermal map and compared with a preset temperature threshold. When the maximum temperature is greater than the preset temperature threshold, the coking pollution level is upgraded.

[0065] It should be noted that the preset temperature threshold is the critical temperature value used in the infrared thermal map to determine the severity of coking. When the temperature in the coking area exceeds the threshold, it indicates that the coking may be melting or the thermal resistance is significant, and the contamination level needs to be upgraded. The preset temperature threshold is set by obtaining historical coking thickness and corresponding infrared temperature data from historical data and establishing a "thickness-temperature" mapping relationship. For example, for every 1mm increase in coking thickness, the wall temperature decreases by 5°C. The temperature corresponding to moderate coking thickness in the historical data is then extracted as the threshold.

[0066] It should be added that the upgrading of the coking pollution level means that when the maximum temperature is greater than a preset temperature threshold, the coking pollution level is upgraded by one level, such as from a medium level to a high level.

[0067] If the area type of a sub-area is dust accumulation, the dust coverage rate of the sub-area is extracted from the surface texture features, and matched with the dust coverage rate intervals corresponding to each dust accumulation level to obtain the dust pollution level of the sub-area.

[0068] It's important to note that the ash coverage range, which measures the ratio of the ash accumulation area to the total area of ​​a sub-region, quantifies the degree of ash contamination and allows for matching different purge intensities, such as high-frequency purges for heavy ash accumulation. The method for setting the ash coverage range for each ash accumulation level is similar to the method for setting the coke thickness range for each coke level described above, so it will not be repeated here.

[0069] The low-temperature coverage area of ​​the sub-region is obtained from the infrared thermal map and compared with a preset low-temperature coverage area threshold. When the low-temperature coverage area is greater than the preset low-temperature coverage area threshold, the dust accumulation pollution level is upgraded.

[0070] It's important to note that the preset low-temperature coverage area threshold is the critical value for the coverage ratio of the low-temperature area to the sub-area in the infrared thermal image. This percentage is used to verify the degree of dust accumulation coverage and avoid misjudgment based on a single coverage ratio, such as when loose dust accumulation is large in area but thin in thickness. The method for obtaining this threshold is similar to the method for obtaining the preset temperature threshold above and will not be repeated here.

[0071] If the area type of a sub-area is mixed, the coking pollution level and ash accumulation pollution level of the sub-area are obtained by similar analysis method according to the coking level and ash accumulation level, and the two are used as the mixed pollution level of the sub-area.

[0072] The maximum temperature and low-temperature coverage area of ​​the sub-area are obtained from the infrared thermal map and compared with the preset values ​​respectively. When the maximum temperature is greater than the preset temperature threshold or the low-temperature coverage area is greater than the preset low-temperature coverage area threshold, the mixed pollution level is upgraded to obtain the pollution level of each sub-area.

[0073] The embodiment of the present invention calculates the coking and ash accumulation levels of the mixed area respectively to obtain a comprehensive pollution level, and corrects the level in combination with the infrared temperature and low-temperature area, thereby effectively solving the current problem of lack of mixed pollution treatment and improving the cleaning efficiency of the mixed pollution area.

[0074] The embodiment of the present invention improves cleaning efficiency by dynamically upgrading the pollution level based on the analysis of coking thickness and dust coverage combined with the temperature characteristics of the infrared thermal map. At the same time, the fusion analysis of infrared temperature data and visual features reduces the misjudgment rate of high-temperature coking and avoids the risk of equipment overheating due to misjudgment of pollution levels.

[0075] The cleaning control module is used to match a cleaning method based on the pollution level and perform cleaning control according to the cleaning method.

[0076] Exemplarily, the cleaning control module includes: when the pollution level is a coking pollution level, matching the coking pollution level with a coking cleaning method corresponding to each pollution level to obtain the coking cleaning method.

[0077] According to the matching method of the coking cleaning method, the dust accumulation cleaning method and the mixed cleaning method are matched in the same way.

[0078] In one specific embodiment, the coke cleaning methods corresponding to the coke pollution level are as follows: Low-level uses pulsed steam purging, with a purging pressure set at 0.5-0.8 MPa and a purging time of 10-15 seconds to avoid excessive purging and damage to the furnace wall. Medium-level uses continuous steam purging, with a pressure raised to 0.8-1.2 MPa, alternating with a mechanical vibrating device, and a purging time extended to 20-30 seconds. High-level uses high-pressure water jet cleaning combined with manual assisted coke removal, while also preemptively lowering the furnace temperature to a safe threshold.

[0079] The corresponding dust cleaning methods for each dust pollution level are: Low-level, regular and scheduled purges are performed using compressed air at a pressure of 0.3-0.5 MPa, covering the dust accumulation area. Medium-level, the purge angle is adjusted according to the location of the dust accumulation, and a rotary soot blower is used, with a purge cycle shortened to 4 hours and a pressure of 0.5-0.7 MPa. High-level, the soot blowers are synchronized and combined with an acoustic soot cleaning device, with a frequency of 20-40 kHz and a purge pressure of 0.7-1.0 MPa. The purge frequency can be increased to once every two hours if necessary.

[0080] Mixed pollution level: Combined with the coking and ash accumulation levels, the higher level pollution is dealt with first. For example, if the coking level is higher than the ash accumulation level, the coking cleaning method is executed first, and then the ash accumulation cleaning method is handled.

[0081] The residual analysis module is used to trigger secondary image acquisition of each sub-area after cleaning is completed, obtain secondary visual images of each sub-area, and perform feature comparison with the reference cleaning state image to obtain each residual area.

[0082] See also Figure 3 As shown, exemplarily, the analysis of each residual region includes: W1, extracting the maximum contour thickness and contour area from the secondary visual image of each sub-region.

[0083] W2. Compare the maximum contour thickness and contour area with a reference thickness threshold and a reference contour area threshold corresponding to the reference cleaning state image, respectively.

[0084] W3. When the maximum contour thickness is greater than the reference thickness threshold or the contour area is greater than the reference contour area threshold, the sub-region is determined to be a residual region, and then each residual region is obtained.

[0085] The embodiment of the present invention locates the residual area by comparing the secondary visual image with the baseline cleaning status image, and triggers secondary cleaning according to the residual parameters, thereby solving the current defect of no verification of the cleaning effect. The cleaning compliance rate is improved through the quantitative evaluation of the residual coke thickness and the ash coverage rate.

[0086] The feedback control terminal is used to determine the residual pollution level based on the residual coke thickness and residual ash coverage of each residual area, and then trigger secondary cleaning in combination with the position coordinates of each residual area and provide corresponding feedback.

[0087] It should be added that the method for determining the residual pollution level is the same as the method for determining the pollution level described above, and will not be repeated here.

[0088] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0089] The above embodiments may be implemented in whole or in part through software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0090] Those skilled in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0091] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0092] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0093] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A boiler furnace cleaning control system based on CCD imaging recognition, characterized by: include: The boiler furnace is divided into sub-areas according to the furnace structure characteristics, and the visual image and infrared thermal map of each sub-area are collected by CCD; Extracting surface texture features from the visual image and dividing each sub-region into a region type; Analyze the coke thickness and ash coverage rate based on the area type of each sub-area, and determine the pollution level of each sub-area by combining infrared thermal map; Matching a cleaning method based on the pollution level and performing cleaning control for the cleaning method; After cleaning is completed, the secondary image acquisition of each sub-area is triggered to obtain the secondary visual image of each sub-area, and the feature comparison is performed with the reference cleaning state image to obtain each residual area; Based on the residual coke thickness and residual ash coverage of each residual area, the residual pollution level is determined, and then combined with the position coordinates of each residual area to trigger secondary cleaning and provide corresponding feedback.

2. The boiler furnace cleaning control system based on CCD imaging recognition according to claim 1, characterized in that: The regional classification of each sub-region includes: Based on the spatial distribution of various structural components in the furnace structure characteristics, the furnace is divided into three layers: upper, middle and lower areas along the vertical direction of the furnace; Each layer area is divided into several sector-shaped areas according to a preset circumferential angle and used as sub-areas.

3. The boiler furnace cleaning control system based on CCD imaging recognition according to claim 1, characterized in that: The regional classification of each sub-region includes: Q1. Determine the contour type of each contour in each sub-region based on the thickness, color and texture of each contour in each sub-region in the surface texture features; Q2. Extracting the area of ​​each contour in each sub-region from the surface texture features, and then combining the contour type to obtain the dust accumulation area and coking area of ​​each sub-region; Q3. Normalize the dust accumulation area and coke accumulation area of ​​each sub-region to obtain the dust accumulation coverage rate and coke accumulation coverage rate of each sub-region; Q4. Determine the area type of each sub-area based on the dust accumulation coverage rate and coking coverage rate of each sub-area.

4. The boiler furnace cleaning control system based on CCD imaging recognition according to claim 3 is characterized in that: The determining of the contour type of each contour in each sub-region includes: Extract the middle value from the thickness range corresponding to the dust accumulation type as the dust accumulation reference thickness; Calculate the similarity between the thickness of each contour and the dust accumulation reference thickness to obtain the dust accumulation thickness similarity of each contour; The dust color similarity and dust texture similarity of each contour are calculated similarly according to the dust thickness similarity calculation method; The dust thickness similarity, dust color similarity and dust texture similarity of each contour are weighted and fused to obtain the comprehensive dust similarity of each contour; The coking comprehensive similarity of each contour is obtained by the same analysis method according to the ash accumulation comprehensive similarity. Compare the comprehensive similarity of dust accumulation and coking of each contour, and select the type corresponding to the high similarity as the initial type of the contour; With the center point of the contour as the center of the circle and the preset radius as the radius, the influence range of each contour is circled, and each contour within the influence range is obtained as each influencing contour. Then, the contour type is secondary verified based on each influencing contour corresponding to each contour to obtain the contour type of each contour in each sub-area.

5. The boiler furnace cleaning control system based on CCD imaging recognition according to claim 4, characterized in that: The analysis of each contour type in each sub-area includes: Based on the initial type of each impact profile, the number of impact profiles consistent with the profile type and the total number of impact profiles are counted, and the ratio of the two is used as the collaborative verification coefficient of the profile; The collaborative verification coefficient is compared with a preset verification threshold. When the collaborative verification coefficient is greater than or equal to the preset verification threshold, the initial type of the contour is used as the contour type of the contour; otherwise, the mixed type is used as the contour type of the contour, thereby obtaining the contour type of each contour in each sub-area.

6. The boiler furnace cleaning control system based on CCD imaging recognition according to claim 3, characterized in that: The analysis of the dust accumulation area and coking area of ​​each sub-region includes: Extracting contours of mixed types from the contour types and recording them as mixed contours; Normalize the comprehensive similarity of dust accumulation and the comprehensive similarity of coking of each mixed profile to obtain the dust accumulation ratio and coking ratio of each mixed profile; The area of ​​each mixed profile is distributed according to the dust accumulation ratio and the coking ratio to obtain the dust accumulation area and the coking area of ​​each mixed profile; According to the allocation method of the dust accumulation area and the coking area of ​​each mixed profile, the areas of the profiles with the profile types of dust accumulation and coking are allocated in the same way, thereby obtaining the dust accumulation area and the coking area of ​​each profile; The dust accumulation area and coking area of ​​each contour in each sub-region are summed up respectively to obtain the dust accumulation area and coking area of ​​each sub-region; The dust accumulation area and coking area of ​​each sub-region are normalized to obtain the dust accumulation coverage rate and coking coverage rate of each sub-region.

7. The boiler furnace cleaning control system based on CCD imaging recognition according to claim 3, characterized in that: Determining the area type of each sub-area includes: If the dust accumulation coverage of a sub-region is greater than or equal to its preset dust accumulation threshold and the coking coverage is less than its preset coking threshold, the region type of the sub-region is determined to be dust accumulation; If the coking coverage of a sub-region is greater than or equal to its preset coking threshold and the dust accumulation coverage is less than its preset dust accumulation threshold, the region type of the sub-region is determined to be coking; If the coking coverage of a sub-region is less than its preset coking threshold and the dust accumulation coverage is less than its preset dust accumulation threshold, the region type of the sub-region is determined to be mixed, and then the region type of each sub-region is obtained.

8. The boiler furnace cleaning control system based on CCD imaging recognition according to claim 1, characterized in that: Determining the pollution level of each sub-area includes: If the area type of a sub-area is coking, the maximum coking thickness is extracted from the surface texture features as the reference coking thickness of the sub-area, and it is matched with the coking thickness interval corresponding to each coking level to obtain the coking pollution level of the sub-area; Obtaining the maximum temperature of the sub-area from the infrared thermal map and comparing it with a preset temperature threshold; when the maximum temperature is greater than the preset temperature threshold, upgrading the coking pollution level; If the area type of a sub-area is dust accumulation, the dust coverage rate of the sub-area is extracted from the surface texture features, and matched with the dust coverage rate interval corresponding to each dust accumulation level to obtain the dust pollution level of the sub-area; Obtaining the low-temperature coverage area of ​​the sub-region from the infrared thermal map, and comparing it with a preset low-temperature coverage area threshold; when the low-temperature coverage area is greater than the preset low-temperature coverage area threshold, upgrading the dust accumulation pollution level; If the area type of a sub-area is mixed, the coking pollution level and ash accumulation pollution level of the sub-area are analyzed in the same way as described above, and the two are used as the mixed pollution level of the sub-area; The maximum temperature and low-temperature coverage area of ​​the sub-area are obtained from the infrared thermal map and compared with the preset values ​​respectively. When the maximum temperature is greater than the preset temperature threshold or the low-temperature coverage area is greater than the preset low-temperature coverage area threshold, the mixed pollution level is upgraded to obtain the pollution level of each sub-area.

9. The boiler furnace cleaning control system based on CCD imaging recognition according to claim 1, characterized in that: The cleaning control includes: When the pollution level is a coking pollution level, the coking pollution level is matched with the coking cleaning method corresponding to each pollution level to obtain the coking cleaning method; According to the matching method of the coking cleaning method, the dust accumulation cleaning method and the mixed cleaning method are matched in the same way.

10. The boiler furnace cleaning control system based on CCD imaging recognition according to claim 1, characterized in that: The analysis of each residual region includes: W1, extract the maximum contour thickness and contour area from the secondary visual image of each sub-region; W2. Compare the maximum contour thickness and contour area with the reference thickness threshold and reference contour area threshold corresponding to the reference cleaning state image respectively; W3. When the maximum contour thickness is greater than the reference thickness threshold or the contour area is greater than the reference contour area threshold, the sub-region is determined to be a residual region, and then each residual region is obtained.

Images