A method for detecting the state of a cutting line of a photovoltaic slicing machine

Abstract

This invention relates to the field of photovoltaic silicon wafer processing technology, and provides a method for detecting the cutting line status of a photovoltaic slicing machine. The method includes: Step 1, a vision acquisition system acquires a first mesh image and a second mesh image respectively; Step 2, an industrial control computer performs jumper analysis on the acquired first mesh image; Step 3, the industrial control computer performs bow analysis on the acquired second mesh image. Step 2 includes: cropping an analysis area to form a jumper analysis image; converting the jumper analysis image into a regular rectangular image; performing image enhancement processing on the converted rectangular image; performing mean filtering on the enhanced image; performing difference processing between the converted rectangular image and the filtered image; filtering feature images based on a brightness threshold; and performing contour matching and searching on the binary image to determine the jumper position. This invention can analyze jumpers and bows, achieving automatic monitoring of the slicing process and improving the efficiency and automation of cutting line status detection.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic silicon wafer processing technology, and in particular to a method for detecting the cutting line status of a photovoltaic slicing machine. Background Technology

[0002] In the photovoltaic silicon wafer industry, the produced ingots ultimately need to be squared and sliced ​​to transform the silicon material into square silicon wafers with a thickness of 0.15mm. The final machining process involves using a slicing machine to cut the square ingots of a certain length into slices.

[0003] like Figures 1-3 As shown, the slicing machine consists of two large wire rollers 1 that drive a wire mesh 2 composed of a row of cutting wires in a reciprocating motion. A square bar 3 feeds the wafer from top to bottom to cut the wafer. Since the silicon wafer is only 0.15mm thick, the diameter of the cutting wires is usually less than 0.05mm. At the same time, the multiple cutting wires are arranged very densely, and the grooves are designed to be very shallow in order to improve efficiency.

[0004] During the cutting process, when the wire roller 1 rotates at high speed, the cutting wire may jump between the wire slots, resulting in two diamond wires in one slot and none in another. This is commonly known as wire skipping. If wire skipping is not detected in time, repeated cutting can cause pulling, leading to wire breakage and loss of both the cut material and the cutting wire.

[0005] During the cutting process, the square bar 3 continuously feeds towards the cutting filament, causing the entire cutting filament to bend. The amount of bending is called the bow. A larger bow results in higher cutting efficiency, but also puts greater stress on the cutting filament, making it more prone to breakage. Conversely, a smaller bow results in lower cutting efficiency, and with the same slicer feed rate, the cutting depth of the square bar 3 may be insufficient, failing to cut through it completely, requiring repeated manual adjustments. Due to slight differences in the manufacturers and batches of the cutting filament, the bow size can vary significantly even with the same tension and cutting depth. Ensuring stable and efficient cutting with a suitable bow size is a challenge faced by slicers.

[0006] Therefore, during the slicing process of a slicer, it is necessary to check the condition of the cutting line, paying particular attention to issues such as skipped cutting lines and wire bowing.

[0007] Currently, the main method for detecting jumper wires is manual inspection. During the cutting process, the machine is stopped, and a person uses a flashlight to illuminate the area and visually inspects the cut wire area. If no jumper wires are observed, the wire mesh color will be relatively uniform, indicating no jumper wires. When jumper wires are present, the wire groove where the jumper occurred will appear empty, and the uneven wire mesh color will be visible to the naked eye, indicating a potential abnormality, requiring manual intervention. The main method for detecting bow wires is still manual visual inspection, which cannot achieve millimeter-level accuracy and cannot accurately reflect the final result.

[0008] In addition, since one person generally monitors multiple machines in the field, the person should monitor 3-4 times in the cutting process (60 minutes for one cutting) in theory, but in practice, the person is often affected by various work and cannot watch in time, so the line jump often causes the network to be disconnected, resulting in relatively large losses; the line bow also needs to be adjusted repeatedly by the person, and the efficiency is low. SUMMARY

[0009] The present application mainly solves the technical problems of the prior art that the cutting line state detection of the slicing machine is performed by manual detection, the detection result is inaccurate, and the detection efficiency is low, and proposes a photovoltaic slicing machine cutting line state detection method, which analyzes the line jump and line bow by image analysis of the collected line net image, realizes automatic monitoring of the slicing process, and improves the cutting line state detection efficiency and automation degree.

[0010] The present application provides a photovoltaic slicing machine cutting line state detection method, comprising the following processes:

[0011] Step 1, a visual acquisition system acquires a first line net image and / or a second line net image;

[0012] Step 2, a work computer analyzes the first line net image for line jump; and / or, Step 3, the work computer analyzes the second line net image for line bow.

[0013] Preferably, the visual acquisition system comprises an image acquisition device and a light source installed together;

[0014] The visual acquisition system has two arrangement modes, one of which is arranged in the front-rear direction of the line roller of the slicing machine, and the other of which is arranged on the side of the line roller of the slicing machine.

[0015] Preferably, in step 1, the image acquisition device is above and the light source is below, and they are projected towards the line roller and the water belt at the same time;

[0016] The image acquisition device can acquire the line net image between the KI regions, and the corresponding background of the KI regions is the KJ region, while the light source is projected to the KF region when it is projected to the KI region, and the FJ region is blocked by the line roller and cannot be projected; the background KJ of the image acquisition device is greater than KF, mainly because the position of the image acquisition device is higher than that of the light source, and the shielding effect of the line roller on the two is different;

[0017] In the image collected by the image collecting device, the line net and the background of the KH region are both relatively bright, and whether there is a jumper on the image is not clear, so the KH region cannot be used for jumper analysis; while in the HI region, the background is dark and the line net is bright, and the collection effect is obviously better than that of the KH region, which is very conducive to image analysis; wherein, L1 represents the upper straight part of the line net, F point represents the intersection point of the center B point of the light source and the line roller extended to the flushing belt, J point represents the intersection point of the center A point of the image collecting device and the line roller extended to the flushing belt, H point represents the intersection point of AF and L1, I point represents the intersection point of AJ and L1, and K point represents the intersection point of the flushing belt and the line net L1.

[0018] Preferably, the step 2 comprises the following steps 201 to 207:

[0019] Step 201, cutting out an analysis region to form a to-be-jumper-analyzed image;

[0020] Step 202, converting the to-be-jumper-analyzed image into a regular rectangular image to obtain a converted rectangular image;

[0021] Step 203, performing image enhancement processing on the converted rectangular image to obtain an enhanced image;

[0022] Step 204, performing mean filtering on the enhanced image to obtain a filtered image;

[0023] Step 205, performing difference processing on the converted rectangular image and the filtered image to obtain a feature image;

[0024] Step 206, screening the feature image according to a brightness threshold to obtain a binary image;

[0025] Step 207, performing contour matching and searching on the binary image to determine the jumper position and complete the analysis.

[0026] Preferably, the step 202 comprises the following process:

[0027] In the to-be-jumper-analyzed image, select the two outermost complete line segments, and through the four end points of the two line segments, use affine transformation of bilinear interpolation to convert the image into a regular rectangular image.

[0028] Preferably, in step 203, the image enhancement processing method is to reduce the gray scale of the highlight area in proportion, and to increase the gray scale of part of the low light area in proportion.

[0029] Preferably, the step 3 comprises the following process:

[0030] A point represents the center point of the image acquisition device, Kc point represents the center of the image acquisition device chip, f represents the focal length of the image acquisition device, K point represents the point of the image acquisition device facing the horizontal cutting line, K' point represents the point of K point moving down after the line bow appears, the line bow analysis needs to calculate the position of K' point; Kc' point represents the corresponding position of K' on the image acquisition device chip; Ka point is a calculation auxiliary point, and the line segment KKa is perpendicular to the line segment K'A; L2 surface represents the plane of the front end surface of the square bar, N represents the intersection point of the vertical line of the image acquisition device center point to L2, and AN represents the distance from the image acquisition device center point to L2;

[0031] Obtaining the position of A point, the position of K point, the position of M point, the position of Kc point, the length of line segment AN, the length of line segment AK, and the focal length f;

[0032] Using the known length of line segment AN, the length of line segment AK, the position of A point, the position of K point, the position of M point, the position of Kc point, the position of Kc' point, and the focal length f, the length of line segment KK' is calculated to obtain the amount of line bow.

[0033] Preferably, the use of the known length of line segment AN, the length of line segment AK, the position of A point, the position of K point, the position of M point, the position of Kc point, the position of Kc' point, and the focal length f to calculate the length of line segment KK' includes the following process:

[0034] Since the lengths of line segment AN and line segment NK are known conditions, according to the Pythagorean theorem, the length of line segment AK can be obtained:

[0035] AK = SQRT (AN*AN+NK*NK);

[0036] According to the similarity of triangles theorem:

[0037] AKa / f = KKa / KcKc' = AK / AKc';

[0038] Therefore, the lengths of line segment KKa and line segment Aka can be obtained.

[0039] Triangle KK'Ka and triangle ANK' share ∠KK'Ka, and both include a right angle, so the two triangles are similar, and according to the similarity theorem:

[0040] K'Ka / (KK'+KN) = KK' / (K'Ka+AKa) = KKa / AN;

[0041] Therefore, K'Ka = (KK'+KN)*KKa / AN;

[0042] At the same time, according to the Pythagorean theorem:

[0043] K'Ka*K'Ka+KKa*KKa=KK'*KK';

[0044] Available:

[0045] (KK' + KN) * KKa / AN * (KK' + KN) * KKa / AN + KKa * KKa = KK' * KK' ;

[0046] (KKa * KKa / (AN * AN) - 1) * KK' * KK' + 2 * KN * KKa / AN + (KKa * KKa / (AN * AN)) * KN * KN + KKa * KKa = 0;

[0047] This equation can be solved to obtain the length of the line segment KK'.

[0048] The photovoltaic slicing machine cutting line state detection method provided by the application can analyze the image collected by the line net image to analyze the jumper and the line bow, realize automatic monitoring of the slicing process, improve the cutting line state detection efficiency and the automation degree, reduce the risk of slicing machine wire breakage, and effectively improve the workshop production efficiency and slicing effect; reduce the working intensity of the operator. The jumper analysis and the line bow analysis have high detection precision and accurate detection results. The visual acquisition system can adopt two arrangement forms and be located at different positions, but can realize detection. BRIEF DESCRIPTION OF DRAWINGS

[0049] Figure 1 is a schematic top view of the relative positions of the main components of the slicing machine;

[0050] Figure 2 is a schematic front view of the relative positions of the main components of the slicing machine;

[0051] Figure 3 is a schematic right view of the relative positions of the main components of the slicing machine;

[0052] Figure 4 is a flowchart of the photovoltaic slicing machine cutting line state detection method provided by the application;

[0053] Figure 5 is a right view of the arrangement mode one of the visual acquisition system provided by the application;

[0054] Figure 6 is a position point schematic diagram of the arrangement mode one of the visual acquisition system provided by the application;

[0055] Figure 7 is a front view of the arrangement mode two of the visual acquisition system provided by the application;

[0056] Figure 8 is a top view of the arrangement mode two of the visual acquisition system provided by the application;

[0057] Figure 9is the implementation flow chart of the jumper wire analysis process provided by the present application;

[0058] Figure 10 is the image state schematic diagram of the jumper wire analysis process provided by the present application;

[0059] Figure 11 is the principle schematic diagram of the wire bow analysis provided by the present application;

[0060] Figure 12 is the mathematical model schematic diagram of the wire bow analysis provided by the present application;

[0061] Figure 13 is the overall running process schematic diagram of the cutting wire state detection using the present application.

[0062] The reference signs are as follows: 1, wire roller; 2, wire net; 3, square stick; 5, slicing machine shell; 6, water flushing belt; 7, image acquisition device; 8, light source; 9, image to be analyzed in the absence of jumper wire; 10, image to be analyzed in the presence of jumper wire; 11, converted rectangular image in the absence of jumper wire; 12, converted rectangular image in the presence of jumper wire. DETAILED DESCRIPTION

[0063] In order to make the technical problems solved by the present application, the technical solutions adopted and the technical effects achieved more clear, the present application will be further described in detail below with reference to the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the present application, but not to limit the present application. In addition, it should be noted that, in order to facilitate the description, only the parts related to the present application are shown in the drawings, but not all the contents.

[0064] As shown in Figure 4 , the photovoltaic slicing machine cutting wire state detection method provided by the embodiments of the present application comprises the following processes:

[0065] Step 1, the visual acquisition system respectively acquires the first wire net image and / or the second wire net image.

[0066] The first wire net image is acquired in the case that the water flushing belt 6 of the slicing machine is opened, and the water flushing belt 6 has a relatively deep color, which can be used as the background of the first wire net image when shooting. The second wire net image is acquired in the case that the water flushing belt 6 of the slicing machine is closed.

[0067] The visual acquisition system comprises the image acquisition device 7 and the light source 8 installed together, which realizes the shooting of the wire net 2. The light source 8 adopts a low-angle strip spotlight.

[0068] The visual acquisition system has two arrangement modes, the arrangement mode one is arranged in the front-rear direction of the wire roller 1 of the slicing machine, and the arrangement mode two is arranged on the side of the wire roller 1 of the slicing machine.

[0069] As Figure 5 shown, visual acquisition system arrangement mode one, specifically is to set up the visual acquisition system in the outside of the microtome shell 5 and is opposite the observation window of the side of the microtome shell 5. The linear roller 1 is arranged in the front and back direction, the light source 8 is bright and the shooting area, the image acquisition device 7 is above the light source 8 and carries out the network image acquisition. Because the microtome is arranged in the front and back direction of the linear roller 1, if it is arranged in the outside of the microtome shell 5, the image acquisition needs the door action of the microtome, and it can influence the detection rhythm; if it is arranged in the inside of the microtome shell 5, the silicon mud and water vapor can cause the pollution of the image acquisition device 7 to the light source 8 when cutting inside. The observation window is usually set in the side of the microtome, and the image acquisition device 7 can be arranged in the light source 8, thereby avoiding the influence.

[0070] As Figure 6 shown, A point represents the center point of the image acquisition device 7, point B represents the center of the light source 8, C point represents the intersection point of the exit angle of the image acquisition device 7 and the water flushing belt 6, D point represents the intersection point of the exit angle of the image acquisition device 7 and the water flushing belt 6, and ∠CAD constitutes the exit angle of the visual acquisition system. E point represents the intersection point of the light source 8 and the water flushing belt 6, G point represents the intersection point of the light source 8 and the water flushing belt 6, and ∠EBG constitutes the exit angle of the light source 8.

[0071] F point represents the intersection point of the light source 8 and the water flushing belt 6, and the position below F point is blocked by the linear roller 1, forming a dark area. J point represents the intersection point of the image acquisition device 7 and the water flushing belt, and the position below J point is blocked by the linear roller 1, and the visual acquisition system cannot be collected. L1 represents the upper straight part of the wire net 2. The intersection point of AF and L1 is H point, and the intersection point of AJ and L1 is I point. The intersection point of the water flushing belt and the wire net L1 is K point.

[0072] The image acquisition device 7 is above, and the light source 8 is below, and they project to the linear roller 1 and the water flushing belt 6. The image acquisition device 7 can collect the wire net image between KI area, and the corresponding background (the water flushing belt 6) of the area is KJ area, and the light source 8 projects to the area KF area, and FJ area is blocked by the linear roller 1 and cannot be projected. The background KJ>KF of the image acquisition device 7, and the main reason is that the position of the image acquisition device 7 is higher than the light source 8, so that the blocking influence area of the linear roller 1 to the two is different.

[0073] In the image collected by the image acquisition device 7, the wire net and the background of KH area are relatively bright, so that whether there is a jumper on the image is not clear, and KH area cannot be used for jumper analysis. In HI area, the background is dark area, and the wire net is bright, so that the collection effect is obviously better than that of KH area, and it is very beneficial to image analysis, and the present application uses HI area for jumper analysis.

[0074] Based on the actual layout of the slicing machine, the light source 8 is used at a low angle to highlight the line net 2, and at the same time, the dark area is formed with the shielding of the line roller 1, and at the same time, the image acquisition device 7 is used at a high angle to collect the image of the line net 2, and the image corresponding to the dark area is analyzed, which is the core feature of the scheme.

[0075] As shown in Figure 7 , 8 , the second arrangement mode of the visual acquisition system is to arrange the visual acquisition system on the side of the line roller 1 of the slicing machine. The light source 8 is arranged on the side of the line roller 1 to highlight the shooting area, and the image acquisition device 7 is arranged above to take a photo. This arrangement is arranged on the side of the slicing machine, but through Figure 7 the top view, it can be seen that the image acquisition device 7 and the light source 8 are inclined towards the shooting area and are still in front of the shooting area in the front-rear direction, and compared with the first arrangement mode, there is little change in the front-rear direction, and it belongs to still in the front-rear direction but with a certain angle of inclination, so the line net images collected by the two arrangement modes can be analyzed for subsequent detection.

[0076] Step 2: The industrial computer analyzes the first line net image collected.

[0077] The first line net image has obtained a relatively good image effect, in order to analyze more accurately, the method based on mathematical model transformation of the present application can not only analyze whether there is a jump line, but also accurately analyze the position of the jump line, so as to provide position information for manual intervention.

[0078] As shown in Figure 9 , step 2 specifically includes the following steps 201 to 207:

[0079] Step 201: Cutting out the analysis area to form the jump line analysis image.

[0080] Since the first line net image collected has bright areas and dark areas, when analyzing the jump line, the dark area which is beneficial to image analysis, i.e. the analysis area, needs to be cut out to form the jump line analysis image (as shown by reference numerals 9 and 10 in Figure 10 ). The jump line analysis image is affected by the distance between the cutting line and the image acquisition device 7, the far end has a large field of view and collects many line bundles, and the near end has a small field of view and collects few line bundles. When there is no jump line, the jump line analysis image is as shown by reference numeral 9 in Figure 10 ; when there is a jump line, the jump line analysis image is as shown by reference numeral 10 in Figure 10 .

[0081] Step 202: Converting the jump line analysis image into a regular rectangular image to obtain the converted rectangular image.

[0082] Specific method can be used, in the image to be analyzed, the two outermost lines of the complete line segment (except the short line segment on both sides) are selected, and the image is converted into a regular rectangular image by using the affine transformation of bilinear interpolation through the four end points of the two line segments. The converted rectangular image is as shown in FIG. 11, 12. Figure 10 When there is no jump line, the converted rectangular image is as shown in FIG. 11; when there is a jump line, the converted rectangular image is as shown in FIG. 12. Figure 10 When there is no jump line, the converted rectangular image is as shown in FIG. 11; when there is a jump line, the converted rectangular image is as shown in FIG. 12. Figure 10 When there is no jump line, the converted rectangular image is as shown in FIG. 11; when there is a jump line, the converted rectangular image is as shown in FIG. 12.

[0083] Step 203, image enhancement processing is performed on the converted rectangular image to obtain an enhanced image.

[0084] In the acquisition process, the effect of the first line net image will be much worse than expected due to the influence of the dense beam, water flushing and cutting of silica mud, so the converted rectangular image needs to be subjected to image enhancement processing.

[0085] The image enhancement processing mainly processes the affected areas on the image. When the adjacent cutting lines appear water film, the acquisition area will be overexposed due to the influence of light. The image enhancement processing method mainly reduces the gray scale ratio of the highlight area and increases the gray scale ratio of part of the low light area.

[0086] By using the floating ratio method, compared with directly increasing and decreasing the gray scale value, the image information can be effectively preserved, the highlight area is obviously decreased, the dark area is not obviously increased, and the image is well averaged.

[0087] The above method needs a prerequisite that when a jump line appears, water film cannot appear in the jump line area. Through actual test, the line distance becomes wider after the jump line appears, and the water film disappears faster, so by controlling the acquisition time after stopping, the situation can be avoided. Therefore, the image enhancement processing process is not affected.

[0088] After the image enhancement processing, a relatively uniform image can be obtained, and when a jump line appears, a dark feature will appear at the corresponding position on the image.

[0089] Step 204, mean filtering is performed on the enhanced image to obtain a filtered image.

[0090] The mean filtering can well eliminate the line features on the image. The mean filtering can be in the form of median filtering, and the filtered image will not have obvious transition.

[0091] Step 205, difference processing is performed on the converted rectangular image and the filtered image to obtain a feature image.

[0092] Since the filtered image no longer has the line-shaped feature of the jump line, the two are differentiated, and the line-shaped feature is first screened out to obtain a feature image. The feature image is an eight-bit gray image of 0-255.

[0093] In step 206, the feature image is screened according to a brightness threshold to obtain a binary image.

[0094] The binary image refers to an image in which the gray scale has only two levels, that is, the gray value of any pixel point in the image is 0 or 255, representing black and white, respectively.

[0095] In step 207, contour matching and searching are performed on the binary image to determine the position of the jump line and complete the analysis.

[0096] In step 3, the industrial computer performs line arch analysis on the collected second wire mesh image.

[0097] Steps 3 and 2 of this embodiment are performed in sequence, or one of them can be selected for detection and analysis according to actual needs.

[0098] A mathematical model as shown in Figure 11 、 12 is established. A point represents the center point of the image acquisition device 7, Kc point represents the center of the chip of the image acquisition device 7, f represents the focal length of the image acquisition device 7, K point represents the point at which the image acquisition device 7 directly faces the horizontal cutting line (i.e., the ideal intersection position of the square bar 3 and the wire mesh 2), K' point represents the point to which the K point is lowered after the line arch appears, and the line arch analysis needs to calculate the position of the K' point (i.e., the actual intersection position of the square bar 3 and the wire mesh 2). Kc' point represents the corresponding position of K' on the chip of the image acquisition device 7. Ka point is a calculation auxiliary point, and the line segment KKa (calculation auxiliary line) is perpendicular to the line segment K'A. L2 plane represents the plane on which the front end surface of the square bar 3 is located, N represents the intersection point of the vertical line from the center point of the image acquisition device 7 to the L2 plane, and AN represents the distance from the center point of the image acquisition device 7 to the L2 plane.

[0099] The image acquisition device 7 directly faces the K point, and after the line arch appears, the K point is lowered to K'. The position of K' needs to be calculated, i.e., the actual intersection position of the square bar 3 and the wire mesh 2. The length of the line segment KK' is the amount of the line arch.

[0100] Since the positions of the image acquisition device 7 and the light source 8 are known when the visual acquisition system is arranged, the relative position of the image acquisition device 7 to the wire roller 1 of the slicing machine is fixed when the image acquisition device 7 is working, and the working position of the square bar 3 is known, the line position analysis can obtain the positions of A point, K point, M point, Kc point, the length of line segment AN, the length of line segment AK, the focal length f, and the position of Kc' point, which is realized by finding the boundary of the second wire mesh image.

[0101] Since Kc is the center of the collection system, the corresponding external position is K point, the K' point position is calculated by Kc' point, and the real line arch value can be obtained, and the specific algorithm used is as follows:

[0102] The length of line segment KK' is calculated by using the known length of line segment AN, the length of line segment AK, the position of point A, the position of point K, the position of point M, the position of point Kc, the position of point Kc', and the focal length f, and the amount of line arch is obtained.

[0103] Line segment KKa is perpendicular to line segment K'A, so triangle AKKa is similar to triangle AKcKc'.

[0104] Since the lengths of line segment AN and line segment NK are known conditions, according to the Pythagorean theorem, the length of line segment AK can be obtained:

[0105] AK = SQRT(AN*AN+NK*NK);

[0106] According to the similarity theorem of triangles:

[0107] AKa / f = KKa / KcKc' = AK / AKc';

[0108] Therefore, the length of line segment KKa and line segment Aka can be obtained.

[0109] Triangle KK'Ka shares ∠KK'Ka with triangle ANK', and both include a right angle, so the two triangles are similar, and according to the similarity theorem, we get:

[0110] K'Ka / (KK'+KN) = KK' / (K'Ka+AKa) = KKa / AN;

[0111] Therefore, K'Ka = (KK'+KN)*KKa / AN;

[0112] At the same time, according to the Pythagorean theorem:

[0113] K'Ka*K'Ka+KKa*KKa = KK'*KK';

[0114] We can get:

[0115] (KK'+KN)*KKa / AN*(KK'+KN)*KKa / AN+KKa*KKa = K K'*K K';

[0116] (KKa*KKa / (AN*AN)-1)*K K'*K K'+2*KN*KKa / AN+(KKa*KKa / (AN*AN))*KN*KN+KKa*KKa = 0;

[0117] By solving the equation, the length of the line segment KK' can be obtained, which is the amount of the line bow.

[0118] The value can also be set by the specification, and when it exceeds the specified value, an alarm is given in time to remind manual intervention.

[0119] This step can detect the intersection position of the square bar 3 and the wire net 2. When the wire net 2 is forced to form a line bow, the intersection position can be obviously seen to move downward, so the specified feature can be obtained according to the second wire net image.

[0120] As shown in Figure 13 The overall running process of the cutting line state detection by using the method is shown in the figure: the slicer runs→preparation for detection→the line roller 1 stops→the vision acquisition system is in place→the light source 8 is turned on→the image acquisition device 7 acquires images→image processing and analysis→the device is restored→the slicer continues to run. If it is analyzed that the cutting line jumps or forms a line bow, the device alarms and processes, and after processing, the slicer continues to run.

[0121] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: the technical solutions recorded in the foregoing embodiments are modified, or part or all of the technical features are replaced, without making the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for detecting a state of a cutting line of a photovoltaic slicing machine, characterized by, The process comprises the following steps: Step 1, a visual acquisition system acquires a first wire mesh image; The visual acquisition system comprises an image acquisition device (7) and a light source (8) installed together; the visual acquisition system has two arrangement modes, mode one is arranged in the front and back direction of the wire roller (1) of the slicing machine, and mode two is arranged on the side of the wire roller (1) of the slicing machine; In step 1, the image acquisition device (7) is above, and the light source (8) is below, while being projected to the wire roller (1) and the water flushing belt (6); the image acquisition device (7) can acquire the wire mesh image between the KI regions, the corresponding background of which is the KJ region, while the light source (8) is projected to the KF region in the region, and the FJ region is blocked by the wire roller (1) and cannot be projected; the background KJ>KF of the image acquisition device (7) is mainly caused by the difference in the shielding effect of the wire roller (1) on the two, because the position of the image acquisition device (7) is higher than that of the light source (8); in the image acquired by the image acquisition device (7), the wire mesh and the background of the KH region are relatively bright, and it is not clear whether there is a jumper in the image, so the KH region cannot be used for jumper analysis; while in the HI region, the background is dark, and the wire mesh is bright, so the acquisition effect is obviously better than that of the KH region, which is very conducive to image analysis; wherein, L1 represents the upper straight part of the wire mesh (2), F represents the intersection point of the center B point of the light source (8) and the wire roller (1) extended to the water flushing belt (6), J represents the intersection point of the center A point of the image acquisition device (7) and the wire roller (1) extended to the water flushing belt, H represents the intersection point of AF and L1, I represents the intersection point of AJ and L1, and K represents the intersection point of the water flushing belt and the wire mesh L1; Step 2, the industrial computer performs jumper analysis on the acquired first wire mesh image; the step 2 comprises the following steps 201 to 207: Step 201, cutting out the analysis region to form a to-be-jumped analysis image; Step 202, converting the to-be-jumped analysis image into a regular rectangular image to obtain a converted rectangular image; Step 203, performing image enhancement processing on the converted rectangular image to obtain an enhanced image; Step 204, performing mean filtering on the enhanced image to obtain a filtered image; Step 205, performing difference processing on the converted rectangular image and the filtered image to obtain a feature image; Step 206, screening the feature image according to the brightness threshold to obtain a binary image; Step 207, performing contour matching and searching on the binary image to determine the jumper position and complete the analysis.

2. The photovoltaic slicer cut line status detection method of claim 1, wherein, The step 202 comprises the following process: In the to-be-jumped analysis image, select the two outermost complete line segments, and through the four end points of the two line segments, use the affine transformation of bilinear interpolation to convert the image into a regular rectangular image.

3. The photovoltaic slicer cut line status detection method of claim 1, wherein, In step 203, the image enhancement processing method is to reduce the gray scale of the highlight area by a certain proportion, and to increase the gray scale of part of the low light area by a certain proportion.

4. A method of detecting the state of a dicing line of a photovoltaic dicer, characterized by, The process comprises the following steps: Step 1, a visual acquisition system acquires a second wire mesh image; The visual acquisition system comprises an image acquisition device (7) and a light source (8) installed together; The visual acquisition system has two arrangement modes, arrangement mode one is arranged in front and back direction of the line roller (1) of the microtome, and arrangement mode two is arranged at the side of the line roller (1) of the microtome; Step 2, the industrial computer performs line arch analysis on the collected second line net image; the step 2 includes the following processes: A mathematical model is established; in the mathematical model, the A point represents the center point of the image acquisition device (7), the Kc point represents the center of the chip of the image acquisition device (7), f represents the focal length of the image acquisition device (7), the K point represents a point of the image acquisition device (7) facing the horizontal cutting line, the K' point represents a point to which the K point is lowered after the line arch appears, the line arch analysis needs to calculate the position of the K' point; the Kc' point represents the corresponding position of the K' point on the chip of the image acquisition device (7); the Ka point is a calculation auxiliary point, the line segment KKa is perpendicular to the line segment K'A; the L2 plane represents the plane on which the front end surface of the square bar (3) is located, N represents the intersection point of the vertical line from the center point of the image acquisition device (7) to the L2 plane, and AN represents the distance from the center point of the image acquisition device (7) to the L2 plane; The positions of the A point, the K point, the M point, the Kc point, the length of the line segment AN, the length of the line segment AK, and the focal length f are obtained. The length of the line segment KK' is calculated by using the known length of the line segment AN, the length of the line segment AK, the position of the A point, the position of the K point, the position of the M point, the position of the Kc point, the position of the Kc' point, and the focal length f, and the amount of the line arch is obtained.

5. The photovoltaic slicer cut line status detection method of claim 4, wherein, The length of the line segment KK' is calculated by using the known length of the line segment AN, the length of the line segment AK, the position of the A point, the position of the K point, the position of the M point, the position of the Kc point, the position of the Kc' point, and the focal length f, and the amount of the line arch is obtained. Since the lengths of the line segment AN and the line segment NK are known conditions, the length of the line segment AK can be obtained according to the Pythagorean theorem: AK = SQRT (AN*AN+NK*NK); According to the similarity of triangles theorem: AKa / f = KKa / KcKc' = AK / AKc'; Therefore, the lengths of the line segment KKa and the line segment Aka are obtained. The triangle KK'Ka and the triangle ANK' share ∠KK'Ka, and both include a right angle, so the two triangles are similar, and according to the similarity theorem: K'Ka / (KK'+KN) = KK' / (K'Ka + AKa) = KKa / AN; Therefore, K'Ka = (KK'+KN)*KKa / AN; At the same time, according to the Pythagorean theorem: K'Ka * K'Ka + KKa * KKa = KK'*KK'; Therefore, the length of the line segment KK' is obtained by solving the equation. ； ​

Images