Furnace chamber temperature measurement window cleaning control method and system

Abstract

The invention relates to the technical field of silicon carbide crystal growth detection, in particular to a furnace chamber temperature measurement window cleaning control method and system. According to the method, the image of the temperature measuring window of the furnace chamber is collected, the image data is processed and analyzed, the shielding rate of the temperature measuring window and the technological parameters in the furnace chamber can be accurately obtained, and then the starting, stopping and running time of temperature measuring window purging equipment in the furnace chamber are controlled in an interlocking mode through a shielding rate threshold value comparison algorithm. The problem that manual judgment is inaccurate and not timely can be effectively solved, and the accuracy of furnace chamber temperature measurement and in-chamber temperature control in the crystal growth process is guaranteed; an image processing result is applied to a crystal growth upper control system, a silicon carbide raw material formula is adjusted in real time according to image analysis data, and the crystal growth quality can be effectively improved; by means of the automatic blowing process, the influence of real-time blowing on crystal growth airflow is reduced, manual operation is reduced, and the system is convenient to overall arrange and install, stable and reliable in operation, high in cleaning efficiency, low in implementation cost and easy to popularize.

Description

Technical Field

[0001] The present invention relates to the technical field of silicon carbide crystal growth detection, and particularly to a cleaning control method and system for a temperature measurement window in a furnace cavity. Background Art

[0002] The physical vapor transport method (PVT) is a process method widely used for silicon carbide crystal growth. It mainly sublimes silicon carbide (SiC) raw materials by controlling temperature, pressure, etc. Among them, temperature is the key object of concern during crystal growth. Generally, the temperature in the growth crucible will reach 2300°C. The quality of crystal growth is directly related to temperature. Currently, most silicon carbide crystal growth devices on the market adopt non-contact infrared temperature measurement methods. Usually, an infrared thermometer is installed above a temperature measurement window equipped with a quartz sheet. During installation, the target is observed through the eyepiece of the thermometer to make the target as close as possible to the center position of the sight. The measured temperature can directly reflect the current process temperature.

[0003] During the crystal growth process, it is necessary to avoid the deposition of volatile substances on the temperature measurement window, which affects the temperature measurement effect. Especially in the temperature control stage, temperature is a very important input parameter. To maintain the stability of temperature measurement, it is necessary to avoid the appearance of volatile substances causing the temperature measurement mirror window to be dirty, which affects the crystal growth process. However, due to the impurities in the powder, after the crucible is heated, they will be transported by the gas flow to the vicinity of the temperature measurement window with the lowest temperature and deposited, resulting in a reduction in the infrared light received by the lens of the infrared thermometer, leading to inaccurate temperature measurement. In severe cases, it affects the quality of crystal growth.

[0004] To solve the problem of the lens being blocked by volatile substances, the prior art one proposes to calculate the deviation empirical value through temperature, power, etc., and then judge whether the infrared temperature measurement value is lower than the deviation value calculated in the previous furnace after the power of the next furnace reaches a certain value. The comparison between the two is used as the basis for judging whether the temperature measurement hole is blocked. However, as the number of times the crucible is used increases, the temperature field characteristics will also change, and at the same time, due to different powders used, the empirical value is not credible; the prior art two proposes to use a purger to purge the lens. Since it is impossible to determine whether there is dirt, the conventional method is to purge when the equipment starts, but long-term purging will affect the internal air flow of the crystal growth furnace; the prior art three conducts manual regular inspections, manually judges the lens blocking situation, and manually starts the purging device. Its cleaning efficiency is low, and it will increase the labor cost. Summary of the Invention

[0005] The technical problem to be solved by the present invention is: aiming at the deficiencies of the prior art, to provide a cleaning control method and system for a temperature measurement window in a furnace cavity, which can accurately judge the purging timing by collecting high-temperature furnace cavity images and using the image processing data for the interlock control of the purging equipment of the temperature measurement window in the crystal growth furnace, improve the cleaning efficiency of the temperature measurement window, and have high stability and low labor cost.

[0006] To solve the above technical problems, the technical solution adopted by the present invention is:

[0007] 1. A cleaning control method for the furnace cavity temperature measurement window

[0008] The present invention provides a cleaning control method for the furnace cavity temperature measurement window, which mainly includes the following steps:

[0009] S101. Turn on the temperature measurement window cleaning control system and perform cyclic detection;

[0010] S102. Perform image acquisition and processing: Use an image shooting device (infrared thermometer or camera equipped with a polarization lens) to acquire an image of the furnace cavity temperature measurement window, and use the OPENCV algorithm to obtain image data and perform image processing;

[0011] S103. Obtain the occlusion rate and furnace cavity process parameters: According to the image processing result, obtain the real-time occlusion rate of the current temperature measurement window, and use the socket method to read the furnace cavity process parameters (including parameters such as the internal temperature of the furnace cavity and the attenuation rate) stored in the upper control system of the furnace cavity, and then send the real-time occlusion rate and furnace cavity process parameters to the lower computer PLC to perform the subsequent purge control steps;

[0012] S104. Determine whether the real-time occlusion rate is greater than a preset threshold one. If so, continue to execute step S105. If not, re-execute step S102;

[0013] S105. Start purging by the purging mechanism: Control the solenoid valve of the purging pipeline to open through the PLC controller to start the purging mechanism to purge the temperature measurement window;

[0014] S106. Determine whether the real-time occlusion rate is greater than a preset threshold two. If so, continue to execute step S107. If not, re-execute step S102;

[0015] S107. Start an audible and visual alarm: Control the audible and visual alarm device to issue an alarm for the occlusion of the temperature measurement window to remind the process personnel through the PLC controller.

[0016] Further, in step S102, the use of the OPENCV algorithm to obtain image data and perform image processing includes the following steps:

[0017] S201. Obtain the furnace cavity image and infrared temperature data;

[0018] S202. Crop the effective area of the image, perform threshold processing on the extracted grayscale image using the AdaptiveThreshold method, and use the findContours method to obtain the contour of the dirty area in the image;

[0019] S203. According to the obtained contour of the dirty area, use the contourArea method to calculate the area a1, a2... a of each contourn and add up all the contour areas to obtain the area A1 of the soiled area;

[0020] S204. Determine the total area A2 of the effective image area, and calculate the occlusion rate m of the temperature measurement window as m = A1 / A2;

[0021] S205. Send the occlusion rate data to the PLC controller.

[0022] Furthermore, in step S103, the method of using the socket method to read the furnace chamber process parameters stored in the upper control system of the furnace chamber includes the following steps:

[0023] S301. Establish a Socket connection between the PLC controller and the upper control system of the furnace chamber;

[0024] S302. The PLC controller reads the corresponding buffer area of the upper control system of the furnace chamber to obtain the stored in-chamber crystal growth attenuation rate and the internal temperature of the furnace chamber; the in-chamber crystal growth attenuation rate is determined by comparing the in-chamber crystal quality corresponding to different time periods.

[0025] Furthermore, in step S105, the purging time T of the purging mechanism is adaptively adjusted according to the current real-time occlusion rate of the temperature measurement window and the in-chamber crystal growth attenuation rate. The specific formula is as follows:

[0026] T = T0 + m×K1 - d×K2

[0027] In the formula, T0 is the preset initial purging time, m is the current real-time occlusion rate of the temperature measurement window, K1 is the conversion coefficient between the occlusion rate and the purging time, d is the in-chamber crystal growth attenuation rate, and K2 is the conversion coefficient between the crystal growth attenuation rate and the purging time.

[0028] When the occlusion rate increases, the purging time is extended to enhance the purging effect; when the occlusion rate decreases, the purging time is reduced to lower the system energy consumption; and when the crystal growth attenuation rate increases, the purging time is correspondingly reduced to reduce the impact of real-time purging on the crystal growth gas flow.

[0029] Furthermore, in step S202, the method of cropping the effective image area specifically includes: calculating the spot size according to the installation distance of the lens of the image capturing device, and determining the center coordinates (center_x, center_y) and radius (radius) of the circular ROI, and extracting the effective area in the image based on the center coordinates and the radius.

[0030] Furthermore, in step S202, the method of performing threshold processing on the extracted grayscale image using the AdaptiveThreshold method includes:

[0031] S401, perform grayscale processing on the image of the effective area and convert it into a single-channel grayscale image;

[0032] S402, determine the pixel threshold of the local neighborhood around each pixel in the image according to the image resolution: calculate the arithmetic mean of the pixels in the local neighborhood and subtract the constant C to obtain the pixel threshold;

[0033] S403, respectively determine whether the pixel value of each pixel is greater than the pixel threshold. If so, set the pixel value of this pixel to 255. If not, set the pixel value of this pixel to 0.

[0034] Further, in step S202, the method of using the findContours method to obtain the contour of the dirty area in the image includes:

[0035] S501, image binarization processing: convert the input image into a single-channel 8-bit binary image;

[0036] S502, set the contour detection mode: select the contour retrieval mode, specify the storage method of the contour points, and control the contour accuracy and data volume;

[0037] S503, execute the findContours contour search function according to the input binary image and the set contour detection mode, and output the contour processing result.

[0038] Further, in step S203, the method of using the contourArea method to calculate the area of each contour includes:

[0039] S601, input the contour processing result data;

[0040] S602, call the contourArea contour area calculation function, and calculate the algebraic area of the closed area based on the coordinates of each contour point through Green's formula;

[0041] S603, output the calculation results a1, a2... a of the area of each contour n 。

[0042] II. A cleaning control system for the furnace cavity temperature measurement window

[0043] Based on the same inventive concept, the present invention also provides a cleaning control system for the furnace cavity temperature measurement window adopting the above-mentioned cleaning control method, which mainly includes:

[0044] 1) Image shooting device: The image shooting device is specifically an infrared thermometer or a camera equipped with a polarization lens;

[0045] 2) Image processing system: The image processing system is equipped with image processing software, which is used to acquire the images collected by the image capturing device and perform image processing and analysis to obtain the temperature measurement window occlusion data;

[0046] 3) Purge module of the PLC controller: It is used to control the corresponding purge action of the purge mechanism according to the instructions from the host computer. The PLC controller is integrally installed in the control cabinet on-site and is connected to the host computer through an Ethernet cable. It also has a logic control module, a communication module, a signal acquisition and processing module, and an alarm module;

[0047] 4) Purge mechanism: The purge mechanism is installed below the temperature measurement window in the furnace cavity. Its purger consists of a circular flange with air holes and an intake pipeline. A on-off solenoid valve is installed on the intake pipeline.

[0048] Among them, the infrared thermometer consists of a lens, a spectral filter, a sensor, a high-definition camera, and a signal processing unit. When in use, it is installed above the temperature measurement window, with a window diameter of 30 mm, the distance between the temperature measurement lens and the surface of the temperature measurement window being greater than 600 mm, and the viewing angle being 90° with the target. The image processing system obtains its video stream signal through Ethernet, uses the image processing software for image acquisition and processing, and transmits the processing result to the PLC controller to control the operation of the purge mechanism;

[0049] The ordinary camera is equipped with a polarization lens to prevent image acquisition failure caused by high temperature and strong light. The polarization lens includes a support frame, a polarizer, and an analyzer. The polarizer is in front of the analyzer, and the analyzer is in front of the camera lens. A reflector is added between the temperature measurement window and the polarizer to meet the installation distance requirements. The video data is transmitted using the RTSP or HTTP protocol of the streaming media protocol.

[0050] The present invention has the following main advantages compared with the prior art:

[0051] 1. The furnace cavity temperature measurement window cleaning control method proposed by the present invention can accurately obtain the occlusion rate of the temperature measurement window and the process parameters in the furnace cavity by collecting the images of the furnace cavity temperature measurement window and processing and analyzing the image data. Then, through the occlusion rate threshold comparison algorithm, it can interlock and control the start-stop and running time of the temperature measurement window purge equipment in the furnace cavity, effectively overcoming the problems of inaccurate and untimely manual judgment, and ensuring the accuracy of furnace cavity temperature measurement and in-cavity temperature control during the crystal growth process;

[0052] 2. The image processing result of the present invention can be used as the basis for subsequent problem positioning. Moreover, by further applying it to furnace cavity abnormal alarm and crystal growth prediction, and cooperating with the upper control system of silicon carbide crystal growth, and adjusting the silicon carbide raw material formula in real time according to the image analysis data, it can effectively improve the crystal growth quality and has broad application prospects;

[0053] 3. The furnace cavity temperature measurement window cleaning control system proposed by the present invention reduces the impact of real-time purging on the crystal growth gas flow through an automated purging process, reduces manual operation, is convenient for overall system layout and installation, operates stably and reliably, has high cleaning efficiency, low implementation cost, and is easy to promote. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a flowchart of the furnace cavity temperature measurement window cleaning control method in the first embodiment of the present invention;

[0055] Figure 2 It is a schematic diagram of the image processing process in the second embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0057] It should be noted that according to the needs of implementation, each step / component described in this application can be split into more steps / components, or two or more steps / components or partial operations of steps / components can be combined into new steps / components to achieve the objectives of the present invention.

[0058] Embodiment 1. This embodiment provides a furnace cavity temperature measurement window cleaning control method, as Figure 1 shown, mainly including the following steps:

[0059] Step S101, turn on the temperature measurement window cleaning control system and perform cyclic detection;

[0060] Step S102, perform image acquisition and processing: obtain data from an infrared thermometer or a camera, and use the OPENCV algorithm for image acquisition and image processing;

[0061] Step S103, obtain the occlusion rate and furnace cavity process parameters:

[0062] Specifically, the occlusion rate is obtained from the image processing result (see Embodiment 2 for details); and the socket method is used to read a specific cache area of the upper control system of silicon carbide crystal growth to obtain information such as the temperature of the pyrometer, the crystal growth attenuation rate, and the internal temperature of the furnace cavity;

[0063] Step S104, set a first threshold to determine whether purging is required;

[0064] Specifically, if the occlusion rate exceeds the first set threshold, the result is transmitted to the PLC; otherwise, it jumps to step 102 to continue image acquisition.

[0065] Step S105, the purging mechanism starts purging.

[0066] Specifically, the purger of the purging mechanism is installed below the temperature measurement window in the vacuum chamber; the programmable logic controller PLC controls the solenoid valve of the purging pipeline to open.

[0067] Step S106, whether the occlusion rate exceeds the second set threshold.

[0068] Specifically, if the threshold exceeds the second threshold, execute step 107 to start the alarm; otherwise, jump to step 102 to continue image acquisition.

[0069] Step S107, audible and visual alarm.

[0070] Through the method of setting two-layer thresholds, not only can occlusions be effectively processed, but also an alarm prompt can be realized when the purging is not clean. In actual application, as a part of the lower-level computer control, this purging control system can be set to start when the overall process begins and end when the process ends.

[0071] Further, the method of obtaining the process parameters in the furnace cavity by using the socket method specifically includes the following steps:

[0072] Step S301, establish a Socket connection between the PLC controller and the upper control system for silicon carbide crystal growth.

[0073] Step S302, the PLC controller reads the corresponding buffer area of the upper control system for silicon carbide crystal growth to obtain the stored in-cavity crystal growth attenuation rate and the furnace cavity internal temperature; the in-cavity crystal growth attenuation rate is determined by comparing the in-cavity crystal quality corresponding to different time periods.

[0074] Further, in step S105, the purging time T of the purging mechanism is adaptively adjusted according to the current real-time occlusion rate of the temperature measurement window and the in-cavity crystal growth attenuation rate. The specific formula is as follows:

[0075] T = T0 + m×K1 - d×K2

[0076] In the formula, T0 is the preset initial purging time, m is the current real-time occlusion rate of the temperature measurement window, K1 is the conversion coefficient between the occlusion rate and the purging time, d is the in-cavity crystal growth attenuation rate, and K2 is the conversion coefficient between the crystal growth attenuation rate and the purging time.

[0077] When the occlusion rate increases, the purging time is extended to enhance the purging effect; when the occlusion rate decreases, the purging time is reduced to lower the system energy consumption; and when the crystal growth attenuation rate increases, the purging time is correspondingly reduced to minimize the impact of real-time purging on the crystal growth gas flow.

[0078] Example 2: An industrial infrared thermometer is used in this example. It has a web server based on the HTTP protocol built-in. By reading a specific address and port number, an MJPEG format image can be obtained. To acquire a clear image, the pyrometer needs to be equipped with a polarizing filter. While collecting the image, it also collects data such as real-time temperature and infrared attenuation rate.

[0079] The specific image processing flow is as Figure 2 shown and mainly includes the following steps:

[0080] Step 201: Obtain the furnace chamber image and infrared temperature.

[0081] Step 202: Crop the effective area, segment the image using a threshold, and obtain the image contour

[0082] Specifically, calculate the spot size based on the installation distance of the infrared lens, determine the center coordinates (center_x, center_y) and radius (radius) of the circular ROI, and extract the ROI based on the center coordinates and mask.

[0083] Furthermore, perform threshold processing on the extracted grayscale image. In this example, since the image illumination is uneven, the AdaptiveThreshold method (adaptive threshold processing method) is used. For each pixel in the image, calculate the threshold of its surrounding blockSize×blockSize neighborhood. According to the adaptiveMethod, calculate the average value or Gaussian weighted sum of this neighborhood, and subtract the constant C from the calculation result to obtain the threshold of this pixel; if the pixel is greater than the threshold, set it to 255, otherwise set it to 0.

[0084] Furthermore, use the findContours method (contour search method) to find the contour edges in the image. These contours represent possible contaminated areas.

[0085] Step 203: For the detected contours, use the contourArea method (contour area calculation method) to calculate the area of each contour a1, a2…a n [ , and sum all the contours to obtain the contaminated area A1.

[0086] Step 204: Determine the total area A2 of the effective area of the image. Divide the contaminated area by the total area to obtain the occlusion rate. The occlusion rate m = A1 / A2.

[0087] Step 205: Transmit the result to the PLC as the basis for system purging and to trigger an alarm. Meanwhile, the upper control system for silicon carbide crystal growth imports the occlusion rate result and the obtained real-time temperature, pressure, attenuation rate, etc. as a training set into the neural network algorithm model to predict crystal growth conditions and notify the process personnel to adjust the silicon carbide raw material formula in real time.

[0088] Further, the AdaptiveThreshold method is used to perform threshold processing on the extracted grayscale image, which specifically includes the following steps:

[0089] Step S401: Grayscale the image of the effective area and convert it into a single-channel grayscale image.

[0090] Step S402: Determine the pixel threshold of the local neighborhood around each pixel in the image according to the image resolution: Calculate the arithmetic mean of the pixels in the local neighborhood and subtract the constant C to obtain the pixel threshold.

[0091] Step S403: Respectively judge whether the pixel value of each pixel is greater than the pixel threshold. If so, set the pixel value of this pixel to 255; if not, set the pixel value of this pixel to 0.

[0092] Further, the findContours method is used to find the contour edges in the image, which specifically includes the following steps:

[0093] Step S501: Binary image processing of the image: Convert the input image into a single-channel 8-bit binary image (such as a black and white image).

[0094] Step S502: Set the contour detection mode: Select the contour retrieval mode (in this embodiment, the outer edge of the object is retrieved), specify the storage method of contour points, and control the contour accuracy and data volume.

[0095] Step S503: Execute the findContours contour search function according to the input binary image and the set contour detection mode, and output the contour processing result.

[0096] Further, the contourArea method is used to calculate the area of each contour, which specifically includes the following steps:

[0097] Step S601: Input the contour processing result data.

[0098] Step S602: Call the contourArea contour area calculation function, and calculate the algebraic area of the closed region based on the coordinates of each contour point through Green's formula (the polygon area calculation formula).

[0099] Step S603: Output the calculation results a1, a2... a of the area of each contour.n 。

[0100] Further, for batch crystal growth equipment, when the equipment working conditions are the same, the aging condition of the crystal growth furnace can be further judged according to different occlusion rates.

[0101] Further, in the case of not using an infrared thermometer, images can be obtained by using an ordinary camera equipped with a polarization lens. The purpose of adding the polarization lens is to prevent strong light interference. The polarizer in the polarization lens can change the light in all directions into one direction, and the purpose of the analyzer is to allow the light of specific light rays to pass through. In order to obtain an ideal image, by adjusting the polarization lens knob multiple times, the light enters the camera at a specific angle, and after obtaining the image, refer to Figure 2 for processing.

[0102] Embodiment 3, based on the same inventive concept, the present invention also provides a cleaning control system for the temperature measurement window of the furnace cavity adopting the cleaning control method as described above, mainly including:

[0103] 1) Image capturing device: The image capturing device is specifically an infrared thermometer or an ordinary camera equipped with a polarization lens;

[0104] The infrared thermometer is composed of a lens, a spectral filter, a sensor, a high-definition camera and a signal processing unit. When in use, it is installed above the temperature measurement window of the crystal growth furnace, the window diameter is 30 mm, the distance between the temperature measurement lens and the surface of the temperature measurement window is greater than 600 mm, the viewing angle is 90° with the target, the video stream signal is obtained by the image processing system through Ethernet, the image is collected and processed by using image processing software, and the processing result is transmitted to the PLC controller to control the operation of the purging mechanism;

[0105] The ordinary camera is equipped with a polarization lens to prevent image capture failure caused by high temperature and strong light. The polarization lens includes a support frame, a polarizer and an analyzer. The polarizer is in front of the analyzer, and the analyzer is in front of the camera lens, and a reflector can be added between the temperature measurement window and the polarizer to meet the installation distance requirement. The video data is transmitted by the streaming media protocol RTSP or the HTTP protocol.

[0106] 2) Image processing system: The image processing system is equipped with image processing software for filtering the collected images, removing noise, obtaining the effective image area, edge extraction, area calculation, etc., to obtain the occlusion situation of the temperature measurement window as the basis for judging whether to start the purging mechanism.

[0107] 3) Purge module of the PLC controller: It is used to control the corresponding purge action of the purge mechanism according to the instructions of the host computer. The PLC controller is integrally installed in the control cabinet at the site of the silicon carbide crystal growth furnace and is connected to the host computer through a network cable. It also has a logic control module, a communication module, a signal acquisition and processing module, and an alarm module;

[0108] 4) Purge mechanism: The purge mechanism is installed below the temperature measurement window in the furnace cavity. Its purger consists of a circular flange with air holes and an air inlet pipeline. A on-off solenoid valve is installed on the air inlet pipeline, and the control of the purge action is achieved by controlling the opening and closing of the solenoid valve.

[0109] Furthermore, the parts not detailed in this application are the same as or implemented using the prior art.

[0110] In summary:

[0111] 1. The furnace cavity temperature measurement window cleaning control method proposed by the present invention can accurately obtain the occlusion rate of the temperature measurement window and the process parameters in the furnace cavity by collecting the image of the furnace cavity temperature measurement window and processing and analyzing the image data. Then, through the occlusion rate threshold comparison algorithm, the start-stop and running time of the temperature measurement window purge equipment in the furnace cavity are interlocked and controlled, which can effectively overcome the problems of inaccurate and untimely manual judgment and ensure the accuracy of furnace cavity temperature measurement and in-cavity temperature control during the crystal growth process;

[0112] 2. The image processing result of the present invention can be used as the basis for subsequent problem positioning. Moreover, by further applying it to furnace cavity abnormal alarm and crystal growth prediction and cooperating with the silicon carbide crystal growth upper control system, and adjusting the silicon carbide raw material formula in real time according to the image analysis data, the crystal growth quality can be effectively improved, and the application prospect is broad;

[0113] 3. The furnace cavity temperature measurement window cleaning control system proposed by the present invention reduces the impact of real-time purging on the crystal growth gas flow through the automated purging process, reduces manual operation, and the overall layout and installation of the system are convenient, the operation is stable and reliable, the cleaning efficiency is high, the implementation cost is low, and it is easy to promote.

[0114] It is easy for those skilled in the art to understand that the above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A cleaning control method for a furnace cavity temperature measuring window, characterized in that, It includes the following steps: S101, Turn on the temperature measurement window cleaning control system and perform cyclic detection; S102, Perform image acquisition and processing: Use an image shooting device to acquire an image of the furnace cavity temperature measurement window, and use a preset image processing algorithm to obtain image data and perform image processing; S103, Obtain the occlusion rate and furnace cavity process parameters: According to the image processing result, obtain the real-time occlusion rate of the current temperature measurement window, and use a preset communication protocol to read the furnace cavity process parameters stored in the upper control system of the furnace cavity; S104, Determine whether the real-time occlusion rate is greater than a preset threshold one. If so, continue to execute step S105. If not, re-execute step S102; S105, The purge mechanism starts purging: Control the solenoid valve of the purge pipeline to open through the PLC controller to start the purge mechanism to purge the temperature measurement window; S106, Determine whether the real-time occlusion rate is greater than a preset threshold two. If so, continue to execute step S107. If not, re-execute step S102; S107, Start the audible and visual alarm: Control the audible and visual alarm device to issue an alarm for the occlusion of the temperature measurement window to remind the process personnel through the PLC controller.

2. The cleaning control method for the furnace cavity temperature measurement window according to claim 1, characterized in that In step S102, the use of a preset image processing algorithm to obtain image data and perform image processing includes the following steps: S201, Obtain the furnace cavity image and infrared temperature data; S202, Crop the effective area of the image, use an adaptive threshold processing algorithm to perform threshold processing on the extracted grayscale image, and use a contour search algorithm to obtain the contour of the dirty area in the image; S203. According to the obtained dirt area contour, use a preset contour area algorithm to calculate the areas a1, a2... a of each contour n , and add up all the contour areas to obtain the dirt area A1; S204, Determine the total area A2 of the effective area of the image, and calculate the occlusion rate m of the temperature measurement window as m = A1 / A2; S205, Send the occlusion rate data to the PLC controller.

3. A method for cleaning and controlling a furnace cavity temperature measurement window according to claim 1, characterized in that In step S103, the use of a preset communication protocol to read the furnace cavity process parameters stored in the upper control system of the furnace cavity includes the following steps: S301, Establish a Socket connection between the PLC controller and the upper control system of the furnace cavity; S302, The PLC controller reads the corresponding buffer area of the upper control system of the furnace cavity to obtain the stored in-cavity crystal growth attenuation rate and the internal temperature of the furnace cavity; The in-cavity crystal growth attenuation rate is determined by comparing the in-cavity crystal quality corresponding to different time periods.

4. The cleaning control method for the furnace chamber temperature measurement window according to claim 2, wherein In step S105, the purge time T of the purge mechanism is adaptively adjusted according to the real-time occlusion rate of the current temperature measurement window and the in-cavity crystal growth attenuation rate. The specific formula is as follows: T = T0 + m×K1 - d×K2 In the formula, T0 is the preset initial purge time, m is the real-time occlusion rate of the current temperature measurement window, K1 is the conversion coefficient between the occlusion rate and the purge time, d is the in-cavity crystal growth attenuation rate, and K2 is the conversion coefficient between the crystal growth attenuation rate and the purge time.

5. The cleaning control method for the furnace cavity temperature measuring window according to claim 2, wherein In step S202, the cropping of the effective area of the image specifically includes: Calculate the spot size according to the installation distance of the lens of the image shooting device, and determine the center coordinates (center_x, center_y) and radius (radius) of the circular ROI, and extract the effective area in the image based on the center coordinates and radius.

6. The cleaning control method for the furnace cavity temperature measurement window according to claim 2, characterized in that In step S202, performing threshold processing on the extracted grayscale image by using an adaptive threshold processing algorithm includes: S401, performing grayscale processing on the image of the valid region and converting it into a single-channel grayscale image; S402, determining the pixel threshold of the local neighborhood around each pixel in the image according to the image resolution: calculating the arithmetic mean of the pixels in the local neighborhood and subtracting a constant C to obtain the pixel threshold; S403, respectively determining whether the pixel value of each pixel is greater than the pixel threshold. If so, setting the pixel value of the pixel to 255; if not, setting the pixel value of the pixel to 0.

7. A method for cleaning and controlling a temperature measurement window of a furnace cavity according to claim 2, characterized in that In step S202, obtaining the contour of the dirty region in the image by using a contour search algorithm includes: S501, performing image binarization processing: converting the input image into a single-channel 8-bit binary image; S502, setting the contour detection mode: selecting the contour retrieval mode, specifying the storage method of the contour points, and controlling the contour accuracy and data volume; S503, executing the findContours contour search function according to the input binary image and the set contour detection mode, and outputting the contour processing result.

8. The cleaning control method for the furnace cavity temperature measuring window according to claim 2, wherein In step S203, calculating the area of each contour by using a preset contour area algorithm includes: S601, inputting the contour processing result data; S602, calling the contourArea contour area calculation function, and calculating the algebraic area of the closed region based on the coordinates of each contour point through Green's formula; S603, output the calculation results a1, a2... a of each contour area n .

9. A cleaning control system for a furnace cavity temperature measuring window adopting the cleaning control method according to any one of claims 1 to 8, characterized in that, including: Image capturing device: The image capturing device is specifically an infrared thermometer or a camera equipped with a polarization lens; Image processing system: The image processing system is equipped with image processing software for obtaining the image collected by the image capturing device and performing image processing and analysis to obtain the temperature measurement window occlusion data; Purge module of the PLC controller: used to control the purge mechanism to perform corresponding purge actions according to the instructions of the upper computer. The PLC controller is integrally installed in the on-site control cabinet, and also has a logic control module, a communication module, a signal acquisition and processing module, and an alarm module; Purge mechanism: The purge mechanism is installed below the temperature measurement window in the furnace cavity. Its purger consists of a circular flange with air holes and an air inlet pipeline, and a on-off solenoid valve is installed on the air inlet pipeline.

10. The furnace cavity temperature measurement window cleaning control system according to claim 9, wherein: The infrared thermometer is composed of a lens, a spectral filter, a sensor, a camera lens and a signal processing unit. When in use, it is installed at a preset height above the temperature measurement window and at a preset angle with the target; the polarization lens installed on the camera includes a support frame, a polarizer and an analyzer. The analyzer is arranged between the polarizer and the camera lens, and a reflector is arranged between the polarizer and the temperature measurement window.

Images