Method and system for flame monitoring and control

Abstract

A method for monitoring a flame of a burner of a lime kiln, comprising: imaging a video stream showing a burner end of the lime kiln; extracting at least one image from the imaged video stream; determining at least one region of interest from the at least one image using a pre-trained algorithm, wherein the at least one region of interest comprises a portion of the at least one image showing a region comprising a flame and / or at least one characteristic portion of the burner end; calculating an area of the at least one characteristic portion based on pixels of the at least one region of interest; and determining at least one quantity of interest based on the calculated area of the at least one characteristic portion.

Description

Technical Field

[0001] This invention relates to a method and arrangement for flame monitoring and control. In particular, but not exclusively, this invention relates to a method and arrangement for monitoring and controlling the flame of a burner in a lime kiln. Background Technology

[0002] White liquor is produced during causticizing treatment, in which green liquor from the recycled boiler reacts with calcium to form white liquor and lime sludge. The following chemical reactions occur:

[0003] Na₂CO₃ + CaO + H₂O → 2NaOH + CaCO₃

[0004] Lime kilns are used in cellulose manufacturing to produce lime from lime sludge for use in preparing white liquor during the cooking process. A calcination reaction occurs:

[0005] CaCO3 → CaO + CO2

[0006] The calcination reaction requires heat, which is generated in a separate burner using oil, gas, or other fuels. The operation of the lime kiln is adjusted by increasing or decreasing the fuel quantity and by controlling the air supply at the burner. The amount of residual calcium carbonate is measured from the produced lime, and its level is kept as constant as possible.

[0007] If too much fuel is used in the burner or there is a malfunction in the air supply, the temperature may rise too high. In this case, the calcium particles will melt, and the kiln lining may be damaged. Again, insufficient fuel will cause a large portion of the lime sludge to remain unreacted, resulting in inadequate causticizing treatment.

[0008] Previously, causticizing treatment was controlled by sampling and measuring the remaining calcium carbonate, which did not provide for real-time adjustments to the treatment. Furthermore, methods using imaging to monitor combustion have been previously disclosed.

[0009] The purpose of this invention is to improve existing solutions by providing a method for monitoring the flame of a burner, so as to monitor causticizing processes and enable easy and rapid control of the process. Summary of the Invention

[0010] Various aspects of the invention are set forth in the claims.

[0011] According to a first exemplary aspect of the present invention, a method for monitoring the flame of a burner in a lime kiln is provided, comprising:

[0012] Image the video stream showing the burner end of the lime kiln;

[0013] Extract at least one image from the imaged video stream;

[0014] At least one region of interest is determined from at least one image using a pre-trained algorithm, wherein the at least one region of interest comprises a portion of at least one image showing a region including at least one characteristic portion of the flame and / or burner end;

[0015] Calculate the area of ​​at least one feature region based on the pixels of at least one region of interest; and

[0016] At least one quantity of interest is determined based on the calculated area of ​​at least one characteristic portion.

[0017] The method may also include comparing the calculated area and / or at least one quantity of interest with a predetermined threshold.

[0018] The method may also include displaying at least one quantity of interest on a user interface component.

[0019] The method may also include adjusting the operation of the burner based on at least one quantity of interest.

[0020] The pre-trained algorithm is created using the following steps:

[0021] Image the video stream showing the burner end of the lime kiln;

[0022] Multiple training images are extracted from the imaged video stream, and each of the multiple training images is segmented into a region of interest, wherein the region of interest of at least one image includes the interior of a lime kiln; and

[0023] The algorithm is trained to identify regions of interest from the segmented image.

[0024] The method may also include:

[0025] Extracting reference images from the captured video stream; and

[0026] The corresponding size of the pixels in the reference image is determined using the known dimensions of the lime kiln visible in the reference image.

[0027] According to a second exemplary aspect of the present invention, a system for monitoring the flame of a burner in a lime kiln is provided, comprising:

[0028] Imaging devices; and

[0029] The processor is configured to perform the method of the first exemplary aspect of the present invention.

[0030] According to a third exemplary aspect of the invention, a computer program including computer-executable program code is provided, which, when executed by a processor, causes the method according to the first exemplary aspect to be performed.

[0031] According to a fourth aspect of the invention, a non-transitory storage medium comprising a computer program of the third aspect is provided.

[0032] The foregoing has described various non-limiting exemplary aspects and embodiments of the invention. The above embodiments are merely illustrative of selected aspects or steps that may be employed in implementing the invention. Some embodiments may be presented with reference only to certain exemplary aspects of the invention. It should be understood that corresponding embodiments may also be applied to other exemplary aspects. Attached Figure Description

[0033] To gain a more complete understanding of exemplary embodiments of the invention, reference is now made to the following description taken in conjunction with the accompanying drawings, in which:

[0034] Figure 1 A flowchart of a method according to an exemplary embodiment of the present invention is shown;

[0035] Figure 2 A further flowchart of a method according to an exemplary embodiment of the present invention is shown; and

[0036] Figure 3 An example image is shown that is used in a method according to an exemplary embodiment of the present invention. Detailed Implementation

[0037] Figure 1 A flowchart of a method according to an exemplary embodiment of the present invention is shown. In one embodiment, the method according to the exemplary embodiment of the present invention is performed by a processor (e.g., a processor of a control system). In one embodiment, the control system is a stand-alone control system configured to control the method, such as a local control system or a cloud-based control system. In a further embodiment, the control system is integrated into a factory-wide control system.

[0038] In step 110, a video stream of the burner end of the lime kiln is imaged. In one embodiment, the imaging apparatus for providing the video stream includes at least one device selected from the group consisting of: a digital camera, a digital still camera configured to capture continuous still images, a high-speed camera, and a thermal imaging device. In one embodiment, the camera is mounted in a suitable location to image a suitable area of ​​the burner end of the lime kiln.

[0039] In step 120, at least one image is extracted from the video stream showing the burner end of the lime kiln. The extracted image, or at least a portion thereof, depicts the burner end of the lime kiln. The image is extracted from the video stream using convenient procedures in image and video processing.

[0040] In step 130, at least one region of interest is determined from at least one image, wherein the at least one region of interest comprises a portion of at least one image showing a region including at least one characteristic portion of the burner end. In one embodiment, the at least one characteristic portion includes a white flame portion, a black flame portion, and / or a lime layer. The determination of the region of interest is performed using an algorithm pre-trained in a manner described in detail below.

[0041] In step 140, the area of ​​at least one feature portion is calculated based on its pixels, according to at least one region of interest that has been determined. In one embodiment, the area of ​​the pixels is determined in SI units during pre-training of the algorithm used. In one embodiment, calculating the area of ​​the at least one feature portion includes calculating a relative area, such as the area of ​​the white flame portion relative to the area of ​​the cross-section of the kiln.

[0042] In step 150, at least one quantity of interest is determined based on the calculated area of ​​at least one feature portion. In one embodiment, the at least one quantity of interest is determined based on the calculated area, location, or further features of one or more feature portions. In one embodiment, the further features of the at least one feature portion include color attributes extracted from the image.

[0043] In one embodiment, at least one quantity of interest is determined directly or using intermediate quantities derived from the calculated area, location, or further features of one or more characteristic portions. For example, in one embodiment, the flame pumping index based on flame length fluctuations is calculated using the maximum and minimum lengths from a specific time period. In yet another embodiment, at least one calculated quantity of interest is selected from the group consisting of: flame angle, black flame area, white flame area, lime area, flame pumping index, kiln dust index, lime overflow area, flame length, flame width, and flame tip angle. In one embodiment, the calculated value of at least one quantity of interest is an absolute or relative value compared to the cross-sectional area of ​​the lime kiln or a portion thereof. In one embodiment, the absolute value includes SI units, pixels, or an index value within a specified range.

[0044] After at least one quantity of interest has been determined, in one embodiment, step 150 compares the result with at least one predetermined threshold for the quantity of interest to determine that the lime kiln is operating in a desired manner. In one embodiment, at least one quantity of interest is sent to a control system. In one embodiment, at least one quantity of interest is displayed on a user interface component such as a display. In one embodiment, the determined at least one quantity of interest is used to adjust or control the combustion process of the lime kiln, for example, by adjusting the air distribution in the burner or by adjusting the fuel quantity.

[0045] Figure 2 A further flowchart is shown according to an exemplary embodiment of the present invention. Figure 2 A method for pre-training an algorithm for determining at least one region of interest is illustrated according to an exemplary embodiment of the present invention. In step 210, a video stream at the burner end of a lime kiln is shown being imaged. In one embodiment, the imaging apparatus for providing the video stream includes at least one device selected from the group consisting of a digital camera, a digital still camera configured to capture continuous still images, a high-speed camera, and a thermal imaging device.

[0046] In step 220, a reference image is extracted from the video stream imaged in step 210. The extracted reference image, or at least a portion thereof, depicts the burner end of a lime kiln. The reference image is extracted from the video stream using convenient procedures in image and video processing. In step 230, the dimensions of the pixels of the reference image, in SI units, are determined using the known dimensions of the burner end of the lime kiln depicted in the reference image. In one embodiment, steps 220 and 230 are skipped if the pixel dimensions have already been determined and / or if the corresponding dimensions in SI units are known.

[0047] In step 240, multiple training images are extracted from the video stream imaged in step 210. The multiple training images are extracted from the video stream using convenient procedures in image and video processing. In one embodiment, the multiple training images include images from different operating scenarios.

[0048] For each of the multiple training images, in step 250, at least one feature portion is segmented into separate images of the multiple training images. In one embodiment, the at least one feature portion includes a white flame portion, a black flame portion, a pipe region, and / or a lime bed. In one embodiment, the segmentation is performed manually, i.e., by a user or operator, for example, by selecting a feature portion from each image.

[0049] In step 260, the algorithm is trained using the multiple training images segmented in step 250 as a training dataset to identify feature regions. In one embodiment, the algorithm includes a deep learning neural network. In a further embodiment, the identification or detection of feature regions is based on further image analysis methods, such as thresholding and morphological operations. In a further embodiment, different methods may be used together or separately.

[0050] Figure 3 An example image 300 is shown, illustrating its use in a method according to an exemplary embodiment of the present invention. Image 300 depicts the burner end of a lime kiln. Figure 3 An example of a characteristic portion of the method according to the invention is shown, and regions of interest 310, 320, and 330 determined from the image showing the characteristic portion. Figure 3 In the example, region of interest 310 includes the white flame portion, region of interest 320 includes the black flame portion, and region of interest 330 includes the lime layer.

[0051] Without limiting the scope, interpretation, or application of the following claims in any way, the technical effect of one or more example embodiments disclosed herein is to provide a monitoring method by which the response to combustion disturbances can be identified and reacted to much faster. Another technical effect of one or more example embodiments disclosed herein is to provide a more stable combustion process by adjusting the fuel and air supply. Yet another technical effect of one or more example embodiments disclosed herein is a more stable calcination process leading to balanced production of white liquor. Still another technical effect of one or more example embodiments disclosed herein is more environmentally friendly operation.

[0052] If necessary, the different functions discussed herein can be performed in different orders and / or simultaneously with each other. Furthermore, if necessary, one or more of the aforementioned functions can be optional or can be combined.

[0053] Although various aspects of the invention are set forth in the independent claims, other aspects of the invention include other combinations of features from the described embodiments and / or dependent claims with features of the independent claims, and not only combinations expressly set forth in the claims.

[0054] It should also be noted that although exemplary embodiments of the invention have been described above, these descriptions should not be considered limiting. Rather, various changes and modifications may be made without departing from the scope of the invention as defined by the appended claims.

Claims

1. A method for monitoring the flame of a burner in a lime kiln, comprising: The video stream is imaged, showing the burner end of the lime kiln; Extract at least one image from the imaged video stream; At least one region of interest is determined from the at least one image using a pre-trained algorithm, wherein the at least one region of interest comprises a portion of the at least one image, and the portion of the at least one image shows a region including at least one characteristic portion of the flame and / or the burner end; The area of ​​the at least one feature portion is calculated based on the pixels of the at least one region of interest; and At least one quantity of interest is determined based on the area calculated from the at least one characteristic portion, wherein the quantity of interest is selected from the group consisting of lime area, kiln dust index, and lime overflow area.

2. The method of claim 1, further comprising comparing the calculated area and / or the at least one quantity of interest with a predetermined threshold.

3. The method according to claim 1 or 2, further comprising displaying the at least one quantity of interest on a user interface component.

4. The method according to claim 1 or 2, further comprising adjusting the operation of the burner based on the at least one quantity of interest.

5. The method according to claim 1, wherein, The pre-trained algorithm is created using the following method: The video stream is imaged, showing the burner end of the lime kiln; Multiple training images are extracted from the imaged video stream, and each of the multiple training images is segmented into a region of interest, wherein the region of interest of at least one image includes the interior of the lime kiln; and The algorithm is trained to identify the region of interest from the segmented image. The segmentation is performed manually.

6. The method according to claim 5, further comprising: Extract a reference image from the imaged video stream; as well as The corresponding size of the pixels in the reference image, in SI units, is determined using the known dimensions of the lime kiln visible in the reference image.

7. A system for monitoring the flame of a burner in a lime kiln, comprising: Imaging devices; as well as A processor configured to perform the method according to any one of claims 1 to 6.

8. A control system for controlling a burner in a lime kiln, comprising the system according to claim 7, wherein, The processor is further configured to adjust the operation of the burner based on the at least one quantity of interest.

9. A computer program product comprising computer-executable program code, which, when executed by a processor, causes to perform the method according to any one of claims 1 to 6.

10. A non-transitory storage medium comprising a computer program product according to claim 9.

Images