Battery piece welding method and system based on infrared lamp tube power control

Abstract

The invention discloses a battery piece welding method and system based on infrared lamp tube power control, and the method comprises the steps: S1, each battery piece is independently placed in a basket, a plurality of baskets are arranged in a matrix form, and a battery matrix is obtained through welding; s2, acquiring an image of the battery matrix, and performing image edge detection; s3, positioning a boundary area of the battery piece, and distinguishing a left edge point and a lower edge point of the battery piece in combination with the image edge obtained in the step S2; s4, geometric features of the lower edge and the left edge of the battery piece are calculated; and S5, the infrared lamp tube is moved to the edge position of the battery piece, the power of the infrared lamp tube is adjusted, and a welding head is driven to conduct welding operation on the edge of the battery piece. According to the invention, an image processing method is adopted and image edge positions are combined to realize accurate positioning of battery piece edges; welding characteristics are analyzed based on geometric characteristics of the edge of the battery piece, so that the power of the infrared lamp tube is accurately adjusted, and welding improvement of the edge of the battery piece is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of battery cell welding, and in particular to a battery cell welding method and system based on infrared lamp power control. Background Art

[0002] With the global population growth and economic development, the demand for energy continues to increase. Solar energy, as an energy source, offers the advantages of abundant resources, clean and pollution-free operation, and widespread distribution, offering enormous application potential. Solar photovoltaic technology has made significant progress. From the initial development of single-crystalline silicon cells to polycrystalline silicon cells, and then to the emergence of new cell technologies such as thin-film cells and perovskite cells, the conversion efficiency of solar cells has continuously improved, from single-digit figures to over 20% today, and even exceeding 40% under laboratory conditions. Simultaneously, the stability and lifespan of cells have also been significantly enhanced. With technological advancements and the expansion of the industry, the cost of solar photovoltaic power generation has continued to decline. Large-scale production, process optimization, and the localization of raw materials have significantly reduced the price of solar cells and modules. Furthermore, the development of system integration technology has reduced the installation and operation and maintenance costs of solar photovoltaic systems, making solar photovoltaic power generation competitive with traditional energy sources in an increasing number of regions.

[0003] Solar cells, as the core components of solar photovoltaic power generation, require reliable connections to transmit current. Cell electrodes are typically made of metal. Good welding creates low-resistance connections between cells and between cells and other conductive components. Low-resistance connections reduce energy loss during current transmission, ensuring smooth current flow from one cell to another, thereby safeguarding the overall performance of the photovoltaic system. Welding not only provides an electrical connection but also mechanically secures the cells. During use, photovoltaic modules are subject to various environmental factors, such as wind, rain, snow, and vibration. Reliable welding ensures that cells are firmly bonded to other components, preventing mechanical stress from causing cell detachment or loosening of connections over long-term use, thereby extending the lifespan of the module. The welds formed during welding protect the cell electrodes and other components from corrosion by moisture, oxygen, and chemicals. High-quality welding forms a dense protective film on the electrode surface, enhancing the cell's corrosion resistance and ensuring that the cell maintains good electrical performance and stability even in harsh environmental conditions.

[0004] In large-scale solar photovoltaic production, increasing production throughput requires accelerated production cycles. However, excessively fast production cycles can lead to reduced welding quality, resulting in increased problems such as inadequate welding and cold joints. Welding process parameters such as temperature, heating rate, and cooling rate must be precisely adjusted based on factors such as cell thickness and the material, shape, and size of the solder ribbon. However, in actual production, due to batch variations in cells and solder ribbons, equipment instability, and other factors, it is difficult to precisely control process parameters within the optimal range for each solder joint, thus affecting consistent welding quality. Welding quality directly impacts the electrical performance matching between cells. Welding defects such as cold joints, missing solder joints, or excessive solder resistance can lead to poor current transfer between cells, resulting in localized hot spots, which can prevent some cells from fully utilizing their power generation capacity and cause power loss in the entire photovoltaic module. According to statistics, poor welding can reduce the power generation efficiency of photovoltaic modules by 5%-15%.

[0005] At the same time, with the continuous development of solar photovoltaic technology, the size of cells is getting smaller and thinner, which places high demands on cell welding inspection technology. Currently, most manufacturers use offline destructive "pull-out testing" to inspect solder joint quality. However, this method cannot effectively detect uncommon defects and inevitably leads to undetected errors, making it difficult to comprehensively and accurately detect welding defects in mass production. Summary of the Invention

[0006] The present invention aims to provide a high-precision cell welding method. This method utilizes image processing to quickly and effectively detect welding quality. Based on precise cell position detection, the power of the infrared lamp is accurately adjusted, and high-precision, high-reliability welding technology is employed to ensure high-quality connections between cells without damaging them.

[0007] The technical solution proposed in the present invention is a cell welding method based on infrared lamp power control, comprising the following steps:

[0008] S1: Each battery cell is placed individually in a basket. Multiple baskets are arranged in a matrix of M rows and N columns. The guide columns of adjacent baskets are welded to achieve welding of adjacent battery cells, and a battery matrix is ​​obtained by welding.

[0009] S2: Acquire an image of the battery matrix, perform image edge detection, obtain a set of image bottom edge points and a set of image left edge points, and then perform straight line fitting to obtain two straight line equations representing the image bottom edge and the image left edge;

[0010] S3: Using the Faster R-CNN model to locate the boundary area of ​​the cell, and combining the two straight line equations obtained in step S2 to distinguish the left edge point of the cell and the bottom edge point of the cell;

[0011] S4: Based on the left edge point and the bottom edge point of the cell, calculate the geometric features of the bottom and left edges of the cell, including the length of the bottom edge of the cell, the angle between the bottom edge of the cell and the horizontal direction, and the grayscale change characteristics of the bottom edge of the cell; the length of the left edge of the cell, the angle between the left edge of the cell and the horizontal direction, and the grayscale change characteristics of the left edge of the cell;

[0012] S5: According to the geometric features of the lower edge and left edge of the battery cell, move the infrared lamp to the edge of the battery cell, adjust the power of the infrared lamp, and drive the welding head to perform welding operations on the edge of the battery cell.

[0013] Optionally, the S2 includes:

[0014] S21: Calculate the grayscale change rate of the image in the row direction and column direction respectively:

[0015]

[0016] Among them, gray hori (x,y) represents the grayscale change rate of the pixel (x,y) in the row direction of the image. verti (x,y) represents the grayscale change rate of the pixel (x,y) in the column direction in the image, Imag represents the collected battery matrix image, Indicates convolution operation, G1 and G2 represent Sobel operators in the horizontal direction and gradient direction respectively, where

[0017]

[0018] S22: Calculate the gradient magnitude of each pixel in the image:

[0019] Gramp(x,y)=((gray hori (x,y)) 2 +(gray verti (x,y)) 2 ) 1 / 2

[0020] Among them, Gramp(x,y) represents the gradient amplitude of the pixel point (x,y);

[0021] Traverse each pixel in the image, compare the gradient amplitude of the pixel with the neighboring pixels, and determine the initial screening edge pixels;

[0022] In the initial screening of pixel edge points, the pixels whose gradient amplitude is greater than the threshold Edge are marked as image edge points, and the image edge point set is obtained:

[0023] Eset0={(x i y i )|i=1,2,…total}

[0024] Among them, Eset0 represents the image edge point set, (x i ,y i ) represents the pixel point (x i) marked as the edge point of the image i ,y i ), total represents the total number of edge points in the image;

[0025] S23: Determine the image lower edge point set Eset bottom , the set of left edge points of the image Eset left :

[0026] Eset bottom ={(x i ,y i )∈Eset0|y i ≤y thresh}

[0027] Eset left ={(x i ,y i )∈Eset0|x i ≤x thresh}

[0028] Among them, y thresh Indicates the height threshold of the lower edge of the image, x thresh Indicates the height threshold of the left edge of the image;

[0029] The least squares method is used to calculate the set Eset bottom , Eset left The points in the image are fitted with straight lines to obtain the equations of two straight lines representing the lower edge and the left edge of the image.

[0030] Optionally, the S22 includes:

[0031] The process of determining the initial screening edge pixels:

[0032] S221: For the pixel point (x, y), compare the gradient amplitude of the pixel point with that of the neighboring pixels. If the following conditions are met at the same time:

[0033]

[0034] Then mark the pixel point (x, y) as the primary screening edge pixel point.

[0035] Optionally, the S23 includes:

[0036] S231: Use the least squares method to calculate the set Eset bottom Fit the points in a straight line and calculate the error function:

[0037]

[0038] Among them, error1 represents the error function, a1 represents the slope of the fitted line, and b1 represents the intercept of the fitted line;

[0039] And solve in It means to find the partial derivative of error1 with respect to a1. It means to find the partial derivative of error1 with respect to b1; find the slope a1 and intercept b1 that minimize the error function error1;

[0040] Get the equation of the line representing the lower edge of the image: y = a1x + b1;

[0041] S232: Using the method of step S231, obtain the equation of the straight line representing the left edge of the image: y=a2x+b2.

[0042] Optionally, the S3 includes:

[0043] S31: Input the image into the Faster R-CNN model, which outputs the classification probability and bounding box regression results for each candidate region. Candidate regions with a classification probability greater than 0.5 are marked as cell boundary regions.

[0044] S32: Based on the pixel points in the cell boundary area, construct a cell edge point set:

[0045] Eset1={(x j y j )|j=1,2,…total′}

[0046] Among them, Eset1 represents the edge point set of the battery cell, (x j ,y j ) represents the edge point of the j-th cell (x j ,y j ), total′ represents the total number of cell edge points;

[0047] For the elements (x j ,y j ), calculate the distance dis from the bottom edge of the image respectively bottom and the distance dis to the left edge of the image left:

[0048]

[0049] If the following equation is satisfied, the element (x j ,y j ) is marked as the lower edge point of the battery cell (m,n):

[0050] (n-1)·Blength-dis0≤dis bottom ≤(n-1)·Blength+dis0

[0051] (m-1)·Bwidth-dis1≤dis left ≤m·Bwidth+dis1

[0052] Where (m,n) represents the index of the cell in the battery matrix, Blength represents the length of each cell in the image, Bwidth represents the width of each cell in the image, dis0 represents the vertical positioning error of the cell, and dis1 represents the horizontal positioning error of the cell.

[0053] If the conditions are met,

[0054] (n-1)·Blength-dis0≤dis bottom ≤n·Blength+dis0

[0055] (m-1)·Bwidth-dis1≤dis left ≤(m-1)·Bwidth+dis1;

[0056] Then the element (x j ,y j ) is marked as the left edge point of the battery cell (m,n).

[0057] Optionally, the S4 includes:

[0058] Calculate the length of the lower edge of the battery cell (m,n) bottom :

[0059] length bottom =x j,max -x j,min ;

[0060] Among them, x j,max Indicates the maximum horizontal coordinate value of the pixel point marked as the lower edge point of the battery cell (m,n), x j,min Indicates the minimum horizontal coordinate value of the pixel point marked as the lower edge point of the battery cell (m,n);

[0061] Calculate the angle between the lower edge of the cell (m,n) and the horizontal direction bottom :

[0062]

[0063] Among them, y j,max Indicates the maximum vertical coordinate value of the pixel point marked as the lower edge point of the battery cell (m,n), y j,min Indicates the minimum vertical coordinate value of the pixel point marked as the lower edge point of the battery cell (m,n);

[0064] Calculate the mean of the gradient amplitude of the pixel marked as the lower edge point of the battery cell (m,n) bottom and variance bottom , used to represent the grayscale change characteristics of the lower edge of the battery cell (m,n);

[0065] Calculate the length of the left edge of the battery cell (m,n) left :

[0066] length left =y j,max -y j,min ;

[0067] Among them, y j,max Indicates the maximum vertical coordinate value of the pixel point marked as the left edge point of the battery cell (m,n), y j,min Indicates the minimum vertical coordinate value of the pixel point marked as the left edge point of the battery cell (m,n);

[0068] Calculate the angle between the left edge of the cell (m,n) and the horizontal direction left :

[0069]

[0070] Among them, x j,max Indicates the maximum horizontal coordinate value of the pixel point marked as the left edge point of the battery cell (m,n), x j,min Indicates the minimum horizontal coordinate value of the pixel point marked as the left edge point of the cell (m,n);

[0071] Calculate the mean of the gradient amplitude of the pixel marked as the left edge point of the cell (m,n) left and variance left , used to represent the grayscale change characteristics of the left edge of the battery cell (m,n).

[0072] Optionally, the S5 includes:

[0073] S51: When the length length of the lower edge of the cell (m,n) bottom does not meet the standard length bottom < Bwidth:

[0074] Then move the infrared lamp tube to [(m - 1)·Bwidth, (n - 1)·Blength], move it at a constant speed in the horizontal direction, set the power of the infrared lamp tube to be always power0, and drive the welding head to perform uniform welding;

[0075] When the length length of the left edge of the cell (m,n) left does not meet the standard length left < Blength:

[0076] Then move the infrared lamp tube to [(m - 1)·Bwidth, (n - 1)·Blength], move it at a constant speed in the vertical direction, set the power of the infrared lamp tube to be always power0, and drive the welding head to perform uniform welding;

[0077] S52: When the angle angle between the lower edge of the cell (m,n) and the horizontal direction bottom does not meet the standard |angle bottom - a1| > angle thre When:

[0078] where angle thre represents the angle deviation threshold;

[0079] Then move the infrared lamp tube to [(m - 1)·Bwidth, (n - 1)·Blength], and move it in the horizontal direction;

[0080] Every other time Δtime, the power of the infrared lamp tube changes from 0 to power0, the infrared lamp tube pauses for 5 seconds without moving, and at this position, the infrared lamp tube drives the welding head for heating and welding; [[ID=*39]]

[0081] When the angle angle between the left edge of the cell (m,n) and the horizontal direction left does not meet the standard |angle left - a3| > angle thre When: Then move the infrared lamp tube to [(m - 1)·Bwidth, (n - 1)·Blength], and move it in the vertical direction;

[0082] Every other time Δtime, the power of the infrared lamp tube changes from 0 to power0, the infrared lamp tube pauses for 5 seconds without moving, and at this position, the infrared lamp tube drives the welding head for heating and welding;

[0083] S53: When the mean value of the gradient amplitude of the pixel point at the lower edge of the cell (m, n) bottom and variance bottom Satisfy|mean bottom -mean std |>mean thre and variance bottom >variance std hour,

[0084] Among them, mean std Indicates the standard value of the mean gradient amplitude, mean thre Indicates the allowable deviation of the mean value of the gradient amplitude, variance std Indicates the standard value of the gradient amplitude variance;

[0085] Then move the infrared lamp to [(m-1)·Bwidth, (n-1)·Blength], and move it horizontally at a constant speed;

[0086] Set the infrared lamp power to power1, where power1 > power0, and use the infrared lamp to dry the soldering material at the bottom edge of the cell to remove the slag. Then set the infrared lamp power to power0 and drive the soldering head to re-solder the guide post at the bottom edge of the cell (m,n) and the guide post at the top edge of the cell (m,n-1).

[0087] When the mean value of the gradient amplitude of the pixel point at the left edge of the cell (m,n) is left and variance left satisfy

[0088] |mean left -mean std |>mean thre and variance left >variance std hour,

[0089] Then move the infrared lamp to [(m-1)·Bwidth, (n-1)·Blength], and move it at a constant speed in the vertical direction;

[0090] Set the infrared lamp power to power1, power1>power0, and use the infrared lamp to dry the welding material at the lower edge of the battery cell to make the slag in the welding material float; then set the infrared lamp power to power0, drive the welding head to re-weld the guide column on the left edge of the battery cell (m,n) and the guide column on the right edge of the battery cell (m,n-1).

[0091] The present invention also provides a cell welding system based on infrared lamp power control, comprising:

[0092] Battery matrix module: Each battery cell is placed in a separate basket, and the baskets are arranged in a matrix. The guide columns of adjacent baskets are welded to form a battery matrix.

[0093] Image edge detection module: collects battery matrix images, calculates the image grayscale change rate, calculates the gradient amplitude of image pixels, compares the gradient amplitude of pixels with that of neighboring pixels, determines the initial screening edge pixels, determines the set of image lower edge points and the set of image left edge points, calculates the error function based on the least squares linear fitting, and constructs a set of equations to solve the linear equations of the lower and left edges of the image respectively;

[0094] Cell edge detection module: identifies the cell boundary area, constructs a set of cell edge points, calculates the distance from the cell edge points to the bottom and left edges of the image, and marks the bottom and left edge points of the cell.

[0095] Cell geometric feature extraction module: calculates the length of the lower edge of the cell, the angle between the lower edge of the cell and the horizontal direction, the grayscale change characteristics of the lower edge of the cell, the length of the left edge of the cell, the angle between the left edge of the cell and the horizontal direction, and the grayscale change characteristics of the left edge of the cell;

[0096] Infrared lamp power control module: identifies the edge features of the battery cell, moves the infrared lamp, adjusts the power of the infrared lamp, and drives the welding head to perform welding operations on the edge of the battery cell.

[0097] Beneficial effects:

[0098] The present invention obtains a battery pack in a desired matrix form by welding the guide posts of adjacent baskets. An image of the battery matrix obtained by welding is captured, and the edges of the battery cells are accurately located through image processing methods combined with the image edge positions. Based on the geometric features of the battery cell edges, welding characteristics are analyzed to quickly and effectively detect welding quality. The power of the infrared lamp is then accurately adjusted based on the accurate position of the battery cell edges and the battery cell welding characteristics, thereby ensuring high-quality connections between the battery cells without damaging the cells.

[0099] During the battery cell welding process, the normal production of battery components is affected due to the occurrence of fragments, cold welds, broken grids, and poor welding offsets. The present invention provides a battery cell welding method for welding quality assessment. The present invention conducts a quality assessment of the weld length of the battery matrix based on the length of the lower edge of the battery cell and the length of the left edge of the battery cell, and adjusts the welding parameters in a timely manner to ensure that the weld length meets the quality requirements. The present invention evaluates the weld direction of the battery matrix based on the angle between the lower edge of the battery cell and the horizontal direction and the angle between the left edge of the battery cell and the horizontal direction, and strengthens it through spot welding to achieve weld direction stability. The present invention evaluates the weld stability of the battery matrix based on the grayscale change characteristics of the lower edge of the battery cell and the grayscale change characteristics of the left edge of the battery cell, and achieves uniform and stable weld thickness by cleaning the weld surface and re-welding.

[0100] The high-precision cell welding method provided by the present invention has an identification rate of defects such as "cold welds", "micro cracks" and "weld ribbon offset" of more than 99%; compared with the traditional "pull-out test", the high-precision cell welding method provided by the present invention effectively improves the welding accuracy; the power generation efficiency of photovoltaic modules manufactured using the cell welding method of the present invention is improved by more than 10% compared with traditional methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0101] Figure 1 A schematic flow chart of a cell welding method based on infrared lamp power control provided by one embodiment of the present invention;

[0102] Figure 2 is the equation of the straight line representing the lower edge of the image obtained in step 2 of the present invention. DETAILED DESCRIPTION

[0103] The present invention will be further described below with reference to the accompanying drawings, but the present invention is not limited in any way. Any changes or substitutions made based on the teachings of the present invention fall within the scope of protection of the present invention.

[0104] Example 1:

[0105] A cell welding method based on infrared lamp power control, such as Figure 1-2 As shown, the following steps are included:

[0106] S1: Each battery cell is placed individually in a basket. Multiple baskets are arranged in a matrix of M rows and N columns. The guide columns of adjacent baskets are welded to achieve welding of adjacent battery cells, and a battery matrix is ​​obtained by welding.

[0107] In an embodiment of the present invention, individual battery cells are placed in a basket, which is composed of a support plate, a load-bearing column, and a guide column. The support plate has a rectangular geometric structure and serves as a supporting platform for the battery cells. Four load-bearing columns are vertically installed at its four corners to support the battery cells. Both the support plate and the load-bearing columns are made of insulating materials, and the gap space formed between the bottom of the battery cell and the support plate serves as a ventilation channel to meet the heat dissipation and ventilation requirements. The guide columns are made of conductive materials, and two guide columns are evenly arranged on each of the four sides of the support plate; the baskets are arranged in a matrix of M rows and N columns according to the pre-planned row and column layout. Through a welding process, the guide columns of adjacent baskets are electrically connected, thereby realizing the electrical welding connection between adjacent battery cells, and finally constructing a battery matrix with an M-row and N-column structure, completing the orderly integration and electrical connection of the battery cells. The specific technical process includes:

[0108] Basket design:

[0109] The rectangular support plate is injection-molded with FRP / PBT composite insulation material (the size is suitable for standard 156mm×156mm or M10 / G12 silicon wafers);

[0110] Glass fiber reinforced nylon load-bearing column arrays are configured at the four corners to form a 2.5±0.1mm bottom ventilation gap;

[0111] Two sets of tinned copper alloy guide posts (Φ3mm×15mm, spacing designed according to a 0.5Voc gradient) are set on each edge;

[0112] Conductive path construction:

[0113] Infrared lamps control the welding head to perform welding;

[0114] Orthogonal matrix welding is performed on adjacent basket guide columns: a series busbar is formed in the row direction, and a bypass diode channel is established in the column direction.

[0115] Soldering joint resistance ≤ 0.5mΩ (compliant with UL 1703 standard)

[0116] Battery Matrix Topology Optimization:

[0117] Using SMT patch technology to pre-install gate line cross compensation circuit and integrate MPPT (maximum power point tracking) compatible 3D-Mesh interconnection architecture;

[0118] Achieve <2% mismatch loss (based on EN 50530 test standard);

[0119] This process utilizes an insulating-conductive composite structure design to achieve a distributed photovoltaic matrix system with adaptive voltage-sharing characteristics while maintaining insulation performance of CTI ≥ 600V. The ohmic contact design of the guide posts significantly reduces series resistance, while the ventilation hole structure keeps the module temperature rise within ΔT < 35K (under STC conditions), achieving an overall conversion efficiency of 22.3% ± 0.5%.

[0120] S2: Collect an image of the battery matrix, perform image edge detection, obtain the image bottom edge point set and the image left edge point set, and then perform straight line fitting to obtain the two straight line equations representing the image bottom edge and the image left edge:

[0121] S21: Calculate the grayscale change rate of the image in the row direction and column direction respectively:

[0122]

[0123] Among them, gray hori (x,y) represents the grayscale change rate of the pixel (x,y) in the row direction of the image. verti (x,y) represents the grayscale change rate of the pixel (x,y) in the column direction in the image, Imag represents the collected battery matrix image, Indicates convolution operation, G1 and G2 represent Sobel operators in the horizontal direction and gradient direction respectively, where

[0124]

[0125] S22: Calculate the gradient magnitude of each pixel in the image:

[0126] Gramp(x,y)=((gray hori (x,y)) 2 +(gray verti (x,y)) 2 ) 1 / 2

[0127] Among them, Gramp(x,y) represents the gradient amplitude of the pixel point (x,y);

[0128] Traverse each pixel in the image, compare the gradient amplitude of the pixel with the neighboring pixels, and determine the initial screening edge pixels;

[0129] In the initial screening of pixel edge points, the pixels whose gradient amplitude is greater than the threshold Edge are marked as image edge points, and the image edge point set is obtained:

[0130] Eset0={(x i y i )|i=1,2,…total}

[0131] Among them, Eset0 represents the image edge point set, (x i ,y i ) represents the pixel point (x i) marked as the edge point of the image i ,y i ), total represents the total number of edge points in the image;

[0132] The process of determining the initial screening edge pixels:

[0133] S221: For the pixel point (x, y), compare the gradient amplitude of the pixel point with that of the neighboring pixels. If the following conditions are met at the same time:

[0134]

[0135] Mark the pixel point (x, y) as the primary screening edge pixel point;

[0136] S23: Determine the image lower edge point set Eset bottom , the set of left edge points of the image Eset left :

[0137] Eset bottom ={(x i ,y i )∈Eset0|y i ≤y thresh}

[0138] Eset left ={(x i ,y i )∈Eset0|x i ≤x thresh}

[0139] Among them, y thresh Indicates the height threshold of the lower edge of the image, x thresh Indicates the height threshold of the left edge of the image;

[0140] The least squares method is used to calculate the set Eset bottom , Eset left Perform straight line fitting on the points in , and obtain the equations of two straight lines representing the lower edge and the left edge of the image:

[0141] S231: Use the least squares method to calculate the set Eset bottom Fit the points in a straight line and calculate the error function:

[0142]

[0143] Among them, error1 represents the error function, a1 represents the slope of the fitted line, and b1 represents the intercept of the fitted line;

[0144] And solve in It means to find the partial derivative of error1 with respect to a1. It means to find the partial derivative of error1 with respect to b1; find the slope a1 and intercept b1 that minimize the error function error1;

[0145] Get the equation of the line representing the lower edge of the image: y = a1x + b1;

[0146] S232: Using the method of step S231, obtain the equation of the line representing the left edge of the image: y=a2x+b2;

[0147] S3: Use the Faster R-CNN model to locate the boundary area of ​​the cell, and combine the two straight line equations obtained in step S2 to distinguish the left edge point and the bottom edge point of the cell:

[0148] S31: Input the image into the Faster R-CNN model, which outputs the classification probability and bounding box regression results for each candidate region. Candidate regions with a classification probability greater than 0.5 are marked as cell boundary regions.

[0149] S32: Based on the pixel points in the cell boundary area, construct a cell edge point set:

[0150] Eset1={(x j y j )|j=1,2,…total′}

[0151] Among them, Eset1 represents the edge point set of the battery cell, (x j ,y j ) represents the edge point of the j-th cell (x j ,y j ), total′ represents the total number of cell edge points;

[0152] For the elements (x j ,y j ), calculate the distance dis from the bottom edge of the image respectively bottom and the distance dis to the left edge of the image left :

[0153]

[0154] If the following equation is satisfied, the element (x j ,y j ) is marked as the lower edge point of the battery cell (m,n):

[0155] (n-1)·Blength-dis0≤dis bottom ≤(n-1)·Blength+dis0

[0156] (m-1)·Bwidth-dis1≤dis left ≤m·Bwidth+dis1

[0157] Where (m,n) represents the index of the cell in the battery matrix, Blength represents the length of each cell in the image, Bwidth represents the width of each cell in the image, dis0 represents the vertical positioning error of the cell, and dis1 represents the horizontal positioning error of the cell.

[0158] If the conditions are met,

[0159] (n-1)·Blength-dis0≤dis bottom ≤n·Blength+dis0

[0160] (m-1)·Bwidth-dis1≤dis left ≤(m-1)·Bwidth+dis1;

[0161] Then the element (x j ,y j ) is marked as the left edge point of the battery cell (m,n).

[0162] S4: Based on the left edge point and the bottom edge point of the cell, calculate the geometric features of the bottom and left edges of the cell, including the length of the bottom edge of the cell, the angle between the bottom edge of the cell and the horizontal direction, and the grayscale change characteristics of the bottom edge of the cell; the length of the left edge of the cell, the angle between the left edge of the cell and the horizontal direction, and the grayscale change characteristics of the left edge of the cell:

[0163] Calculate the length of the lower edge of the battery cell (m,n) bottom :

[0164] length bottom =x j,max -x j,min ;

[0165] Among them, x j,max Indicates the maximum horizontal coordinate value of the pixel point marked as the lower edge point of the battery cell (m,n), x j,min Indicates the minimum horizontal coordinate value of the pixel point marked as the lower edge point of the battery cell (m,n);

[0166] Calculate the angle between the lower edge of the cell (m,n) and the horizontal direction bottom :

[0167]

[0168] Among them, y j,max Indicates the maximum vertical coordinate value of the pixel point marked as the lower edge point of the battery cell (m,n), y j,min Indicates the minimum vertical coordinate value of the pixel point marked as the lower edge point of the battery cell (m,n);

[0169] Calculate the mean of the gradient amplitude of the pixel marked as the lower edge point of the battery cell (m,n) bottom and variance bottom , used to represent the grayscale change characteristics of the lower edge of the battery cell (m,n);

[0170] Calculate the length of the left edge of the battery cell (m,n) left :

[0171] length left =y j,max -y j,min ;

[0172] Among them, y j,max Indicates the maximum vertical coordinate value of the pixel point marked as the left edge point of the battery cell (m,n), y j,min Indicates the minimum vertical coordinate value of the pixel point marked as the left edge point of the battery cell (m,n);

[0173] Calculate the angle between the left edge of the cell (m,n) and the horizontal direction left :

[0174]

[0175] Among them, x j,max Indicates the maximum horizontal coordinate value of the pixel point marked as the left edge point of the battery cell (m,n), x j,min Indicates the minimum horizontal coordinate value of the pixel point marked as the left edge point of the cell (m,n);

[0176] Calculate the mean of the gradient amplitude of the pixel marked as the left edge point of the cell (m,n) left and variance left , used to represent the grayscale change characteristics of the left edge of the battery cell (m,n);

[0177] S5: According to the geometric features of the lower and left edges of the cell, move the infrared lamp to the edge of the cell, adjust the power of the infrared lamp, and drive the welding head to perform welding operations on the edge of the cell:

[0178] S51: When the length length of the lower edge of the solar cell (m,n) bottom does not meet the standard length bottom < Bwidth:

[0179] Then move the infrared lamp tube to [(m - 1)·Bwidth, (n - 1)·Blength], move it uniformly in the horizontal direction, set the power of the infrared lamp tube to be always power0, and drive the welding head to perform uniform soldering;

[0180] When the length length of the left edge of the solar cell (m,n) left does not meet the standard length left < Blength:

[0181] Then move the infrared lamp tube to [(m - 1)·Bwidth, (n - 1)·Blength], move it uniformly in the vertical direction, set the power of the infrared lamp tube to be always power0, and drive the welding head to perform uniform soldering;

[0182] S52: When the angle angle between the lower edge of the solar cell (m,n) and the horizontal direction bottom does not meet the standard |angle bottom - a1| > angle thre :

[0183] where angle thre represents the angle deviation threshold;

[0184] Then move the infrared lamp tube to [(m - 1)·Bwidth, (n - 1)·Blength], and move it in the horizontal direction;

[0185] Every Δtime, the power of the infrared lamp tube changes from 0 to power0, the infrared lamp tube pauses for 5 seconds without moving, and at this position, the infrared lamp tube drives the welding head to perform heating and soldering;

[0186] When the angle angle between the left edge of the solar cell (m,n) and the horizontal direction left does not meet the standard |angle left - a3| > angle thre : Then move the infrared lamp tube to [(m - 1)·Bwidth, (n - 1)·Blength], and move it in the vertical direction;

[0187] Every Δtime, the power of the infrared lamp tube changes from 0 to power0, the infrared lamp tube pauses for 5 seconds without moving, and at this position, the infrared lamp tube drives the welding head to perform heating and soldering;

[0188] S53: When the mean value of the gradient amplitude of the pixel point at the lower edge of the cell (m, n) bottom and variance bottom Satisfy|mean bottom -mean std |>mean thre and variance bottom >variance std hour,

[0189] Among them, mean std Indicates the standard value of the mean gradient amplitude, mean thre Indicates the allowable deviation of the mean value of the gradient amplitude, variance std Indicates the standard value of the gradient amplitude variance;

[0190] Then move the infrared lamp to [(m-1)·Bwidth, (n-1)·Blength], and move it horizontally at a constant speed;

[0191] Set the infrared lamp power to power1, where power1 > power0, and use the infrared lamp to dry the soldering material at the bottom edge of the cell to remove the slag. Then set the infrared lamp power to power0 and drive the soldering head to re-solder the guide post at the bottom edge of the cell (m,n) and the guide post at the top edge of the cell (m,n-1).

[0192] When the mean value of the gradient amplitude of the pixel point at the left edge of the cell (m,n) is left and variance left satisfy

[0193] |mean left -mean std |>mean thre and variance left >variance std hour,

[0194] Then move the infrared lamp to [(m-1)·Bwidth, (n-1)·Blength], and move it at a constant speed in the vertical direction;

[0195] Set the infrared lamp power to power1, power1>power0, and use the infrared lamp to dry the welding material at the lower edge of the battery cell to make the slag in the welding material float; then set the infrared lamp power to power0, drive the welding head to re-weld the guide column on the left edge of the battery cell (m,n) and the guide column on the right edge of the battery cell (m,n-1).

[0196] In the embodiments of the present invention, infrared lamps are used in the solar photovoltaic cell welding process. This process primarily involves using electrical energy to excite the heating element within the lamp to generate infrared radiation. This radiation can directly penetrate the air and act on the surface of the material being welded, converting light energy into heat energy and achieving rapid local heating. Compared to traditional resistance heating, infrared heating offers advantages such as fast response, high thermal efficiency, and uniform temperature distribution. It is particularly suitable for ultra-thin cells (thickness as low as 120-150μm) that are sensitive to welding temperatures. In the embodiments of the present invention, the infrared lamp is precisely positioned to the edge of the cell for heating based on cell edge positioning. Local temperature control is achieved by controlling the power of the infrared lamp, thereby enabling drying, additional welding, spot welding, and re-welding of the welding material.

[0197] In the embodiment of the present invention, Figure 2 This means that after the battery matrix image is processed according to step S2, a set of image lower edge points is obtained, and a straight line equation representing the image lower edge is obtained by fitting. As can be seen from the figure, the fitted straight line includes most of the image lower edge points.

[0198] In an embodiment of the present invention, according to the geometric characteristics of the edge of the battery cell, the infrared lamp is moved to the edge of the battery cell, the power of the infrared lamp is adjusted, and the welding head is driven to perform welding operations on the edge of the battery cell; the length of the weld is a basic geometric characteristic, and too fast welding speed and too small welding current will result in insufficient weld length. At this time, it is necessary to adjust the welding parameters, reduce the welding speed, and appropriately increase the welding current; the angle between the weld and the horizontal direction is used to describe the degree of bending of the weld, and when the welding head moves and deviates, the angle between the weld and the horizontal direction will be offset; at this time, it is necessary to perform additional welding at a certain distance, and use point welding to adjust the edge shape to reduce the offset of the angle between the weld and the horizontal direction; the grayscale change characteristics of the weld are used to describe the smoothness of the welding. If the slag is not completely floated out and the weld is not thoroughly cleaned during the welding process, the grayscale of the weld will change dramatically. At this time, it is necessary to clean the weld surface, increase the welding current, and increase the molten pool temperature to make the slag fully float out, and then re-weld.

[0199] Example 2: The present invention also provides a cell welding system based on infrared lamp power control, comprising the following five modules:

[0200] Battery matrix module: Each battery cell is placed in a separate basket, and the baskets are arranged in a matrix. The guide columns of adjacent baskets are welded to form a battery matrix.

[0201] Image edge detection module: collects battery matrix images, calculates the image grayscale change rate, calculates the gradient amplitude of image pixels, compares the gradient amplitude of pixels with that of neighboring pixels, determines the initial screening edge pixels, determines the set of image lower edge points and the set of image left edge points, calculates the error function based on the least squares linear fitting, and constructs a set of equations to solve the linear equations of the lower and left edges of the image respectively;

[0202] Cell edge detection module: identifies the cell boundary area, constructs a set of cell edge points, calculates the distance from the cell edge points to the bottom and left edges of the image, and marks the bottom and left edge points of the cell.

[0203] Cell geometric feature extraction module: calculates the length of the lower edge of the cell, the angle between the lower edge of the cell and the horizontal direction, the grayscale change characteristics of the lower edge of the cell, the length of the left edge of the cell, the angle between the left edge of the cell and the horizontal direction, and the grayscale change characteristics of the left edge of the cell;

[0204] Infrared lamp power control module: identifies the edge features of the battery cell, moves the infrared lamp, adjusts the power of the infrared lamp, and drives the welding head to perform welding operations on the edge of the battery cell.

[0205] It should be noted that the serial numbers of the above-mentioned embodiments of the present invention are for descriptive purposes only and do not represent the advantages or disadvantages of the embodiments. In addition, the terms "including", "comprising" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "including a ..." does not exclude the presence of other identical elements in the process, device, article or method comprising the element.

[0206] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better embodiment. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present invention.

[0207] The above are only preferred embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A cell welding method based on infrared lamp power control, characterized in that: include: S1: Each battery cell is placed individually in a basket. Multiple baskets are arranged in a matrix of M rows and N columns. The guide columns of adjacent baskets are welded to achieve welding of adjacent battery cells, and a battery matrix is ​​obtained by welding. S2: Acquire an image of the battery matrix, perform image edge detection, obtain a set of image bottom edge points and a set of image left edge points, and then perform straight line fitting to obtain two straight line equations representing the image bottom edge and the image left edge; S3: Using the Faster R-CNN model to locate the boundary area of ​​the cell, and combining the two straight line equations obtained in step S2 to distinguish the left edge point of the cell and the bottom edge point of the cell; S4: Based on the left edge point and the bottom edge point of the cell, calculate the geometric features of the bottom edge and the left edge of the cell, including the length of the bottom edge of the cell, the angle between the bottom edge of the cell and the horizontal direction, and the grayscale change characteristics of the bottom edge of the cell; The length of the left edge of the cell, the angle between the left edge of the cell and the horizontal direction, and the grayscale change characteristics of the left edge of the cell; S5: According to the geometric features of the lower edge and left edge of the battery cell, move the infrared lamp to the edge of the battery cell, adjust the power of the infrared lamp, and drive the welding head to perform welding operations on the edge of the battery cell.

2. The battery cell welding method based on infrared lamp power control according to claim 1, characterized in that: The step S2 comprises: S21: Calculate the grayscale change rate of the image in the row direction and column direction respectively: Among them, gray hori (x,y) represents the grayscale change rate of the pixel (x,y) in the row direction of the image. verti (x,y) represents the grayscale change rate of the pixel (x,y) in the column direction in the image, Imag represents the collected battery matrix image, Indicates convolution operation, G1 and G2 represent Sobel operators in the horizontal direction and gradient direction respectively, where S22: Calculate the gradient magnitude of each pixel in the image: Gramp(x,y)=((gray hori (x,y)) 2 +(gray verti (x,y)) 2 ) 1 / 2 Among them, Gramp(x,y) represents the gradient amplitude of the pixel point (x,y); Traverse each pixel in the image, compare the gradient amplitude of the pixel with the neighboring pixels, and determine the initial screening edge pixels; In the initial screening of pixel edge points, the pixels whose gradient amplitude is greater than the threshold Edge are marked as image edge points, and the image edge point set is obtained: Eset0={(x i y i )|i=1,2,…total} Among them, Eset0 represents the image edge point set, (x i ,y i ) represents the pixel point (x i) marked as the edge point of the image i ,y i ), total represents the total number of edge points in the image; S23: Determine the image lower edge point set Eset bottom , the set of left edge points of the image Eset left : Eset bottom ={(x i ,and i )∈Eset0|y i ≤y thresh } Eset left ={(x i ,and i )∈Eset0|x i ≤x thresh } Among them, y thresh Indicates the height threshold of the lower edge of the image, x thresh Indicates the height threshold of the left edge of the image; The least squares method is used to calculate the set Eset bottom , Eset left The points in the image are fitted with straight lines to obtain the equations of two straight lines representing the lower edge and the left edge of the image.

3. The battery cell welding method based on infrared lamp power control according to claim 2, characterized in that: The step S22 includes: The process of determining the initial screening edge pixels: S221: For the pixel point (x, y), compare the gradient amplitude of the pixel point with that of the neighboring pixels. If the following conditions are met at the same time: Then mark the pixel point (x, y) as the primary screening edge pixel point.

4. The battery cell welding method based on infrared lamp power control according to claim 2, characterized in that: The step S23 includes: S231: Use the least squares method to calculate the set Eset bottom Fit the points in a straight line and calculate the error function: Among them, error1 represents the error function, a1 represents the slope of the fitted line, and b1 represents the intercept of the fitted line; And solve in It means to find the partial derivative of error1 with respect to a1. It means to find the partial derivative of error1 with respect to b1; find the slope a1 and intercept b1 that minimize the error function error1; Get the equation of the line representing the lower edge of the image: y = a1x + b1; S232: Using the method of step S231, obtain the equation of the straight line representing the left edge of the image: y=a2x+b2.

5. The battery cell welding method based on infrared lamp power control according to claim 2, characterized in that: The step S3 comprises: S31: Input the image into the Faster R-CNN model, which outputs the classification probability and bounding box regression results for each candidate region. Candidate regions with a classification probability greater than 0.5 are marked as cell boundary regions. S32: Based on the pixel points in the cell boundary area, construct a cell edge point set: Eset1={(x j and j )|j=1,2,…total′} Among them, Eset1 represents the edge point set of the battery cell, (x j ,y j ) represents the edge point of the j-th cell (x j ,y j ), total′ represents the total number of cell edge points; For the elements (x j ,y j ), calculate the distance dis from the bottom edge of the image respectively bottom and the distance dis to the left edge of the image left : If the following equation is satisfied, the element (x j ,y j ) is marked as the lower edge point of the battery cell (m,n): (n-1)·Blength-dis0≤dis bottom ≤(n-1)·Blength+dis0 (m-1)·Bwidth-dis1≤dis left ≤m·Bwidth+dis1 Where (m,n) represents the index of the cell in the battery matrix, Blength represents the length of each cell in the image, Bwidth represents the width of each cell in the image, dis0 represents the vertical positioning error of the cell, and dis1 represents the horizontal positioning error of the cell. If the conditions are met, (n-1)·Blength-dis0≤dis bottom ≤n·Blength+dis0 (m-1)·Bwidth-dis1≤dis left ≤(m-1)·Bwidth+dis1; Then the element (x j ,y j ) is marked as the left edge point of the battery cell (m,n).

6. The battery cell welding method based on infrared lamp power control according to claim 5, characterized in that: The step S4 comprises: Calculate the length of the lower edge of the battery cell (m,n) bottom : length bottom =x j,max -x j,min ; Among them, x j,max Indicates the maximum horizontal coordinate value of the pixel point marked as the lower edge point of the battery cell (m,n), x j,min Indicates the minimum horizontal coordinate value of the pixel point marked as the lower edge point of the battery cell (m,n); Calculate the angle between the lower edge of the cell (m,n) and the horizontal direction bottom : Among them, y j,max Indicates the maximum vertical coordinate value of the pixel point marked as the lower edge point of the battery cell (m,n), y j,min Indicates the minimum vertical coordinate value of the pixel point marked as the lower edge point of the battery cell (m,n); Calculate the mean of the gradient amplitude of the pixel marked as the lower edge point of the battery cell (m,n) bottom and variance bottom , used to represent the grayscale change characteristics of the lower edge of the battery cell (m,n); Calculate the length of the left edge of the battery cell (m,n) left : length left =y j,max -y j,min ; Among them, y j,max Indicates the maximum vertical coordinate value of the pixel point marked as the left edge point of the battery cell (m,n), y j,min Indicates the minimum vertical coordinate value of the pixel point marked as the left edge point of the battery cell (m,n); Calculate the angle between the left edge of the cell (m,n) and the horizontal direction left : Among them, x j,max Indicates the maximum horizontal coordinate value of the pixel point marked as the left edge point of the battery cell (m,n), x j,min Indicates the minimum horizontal coordinate value of the pixel point marked as the left edge point of the cell (m,n); Calculate the mean of the gradient amplitude of the pixel marked as the left edge point of the cell (m,n) left and variance left , used to represent the grayscale change characteristics of the left edge of the battery cell (m,n).

7. The battery cell welding method based on infrared lamp power control according to claim 6, characterized in that: The S5 includes: S51: When the length of the lower edge of the cell (m,n) bottom does not meet the standard length bottom < Bwidth: Then move the infrared lamp to [(m-1)·Bwidth, (n-1)·Blength], move it horizontally at a uniform speed, set the infrared lamp power to always be power0, and drive the welding head to weld evenly; When the length of the left edge of the battery cell (m,n) left does not meet the standard length left When <Blength: Then move the infrared lamp to [(m-1)·Bwidth, (n-1)·Blength], move it vertically at a constant speed, set the infrared lamp power to always be power0, and drive the welding head to weld evenly; S52: When the angle between the lower edge of the cell (m,n) and the horizontal direction is bottom Does not meet the standard |angle bottom -a1|>angle thre hour: Among them, angle thre Indicates the angle deviation threshold; Then move the infrared lamp to [(m-1)·Bwidth, (n-1)·Blength] in the horizontal direction; Every time Δtime, the power of the infrared lamp changes from 0 to power0, and the infrared lamp pauses for 5 seconds without moving. At this position, the infrared lamp drives the welding head to heat and weld; When the angle between the left edge of the cell (m,n) and the horizontal direction is left Does not meet the standard |angle left -a3|>angle thre When: move the infrared lamp to [(m-1)·Bwidth, (n-1)·Blength], and move in the vertical direction; Every time Δtime, the power of the infrared lamp changes from 0 to power0, and the infrared lamp pauses for 5 seconds without moving. At this position, the infrared lamp drives the welding head to heat and weld; S53: When the mean value of the gradient amplitude of the pixel point at the lower edge of the cell (m, n) bottom and variance bottom Satisfy|mean bottom -mean std |>mean thre and variance bottom >variance std hour, Among them, mean std Indicates the standard value of the mean gradient amplitude, mean thre Indicates the allowable deviation of the mean value of the gradient amplitude, variance std Indicates the standard value of the gradient amplitude variance; Then move the infrared lamp to [(m-1)·Bwidth, (n-1)·Blength], and move it horizontally at a constant speed; Set the infrared lamp power to power1, where power1 > power0, and use the infrared lamp to dry the soldering material at the bottom edge of the cell to remove the slag. Then set the infrared lamp power to power0 and drive the soldering head to re-solder the guide post at the bottom edge of the cell (m,n) and the guide post at the top edge of the cell (m,n-1). When the mean value of the gradient amplitude of the pixel point at the left edge of the cell (m,n) is left and variance left satisfy |mean left -mean std |>mean thre and variance left >variance std hour, Then move the infrared lamp to [(m-1)·Bwidth, (n-1)·Blength], and move it at a constant speed in the vertical direction; Set the infrared lamp power to power1, power1>power0, and use the infrared lamp to dry the welding material at the lower edge of the battery cell to make the slag in the welding material float; then set the infrared lamp power to power0, drive the welding head to re-weld the guide column on the left edge of the battery cell (m,n) and the guide column on the right edge of the battery cell (m,n-1).

8. A cell welding system based on infrared lamp power control, characterized in that: include: Battery matrix module: Each battery cell is placed in a separate basket, and the baskets are arranged in a matrix. The guide columns of adjacent baskets are welded to form a battery matrix. Image edge detection module: collects battery matrix images, calculates the image grayscale change rate, calculates the gradient amplitude of image pixels, compares the gradient amplitude of pixels with that of neighboring pixels, determines the initial screening edge pixels, determines the set of image lower edge points and the set of image left edge points, calculates the error function based on the least squares linear fitting, and constructs a set of equations to solve the linear equations of the lower and left edges of the image respectively; Cell edge detection module: identifies the cell boundary area, constructs a set of cell edge points, calculates the distance from the cell edge points to the bottom and left edges of the image, and marks the bottom and left edge points of the cell. Cell geometric feature extraction module: calculates the length of the lower edge of the cell, the angle between the lower edge of the cell and the horizontal direction, the grayscale change characteristics of the lower edge of the cell, the length of the left edge of the cell, the angle between the left edge of the cell and the horizontal direction, and the grayscale change characteristics of the left edge of the cell; Infrared lamp power control module: identifies the edge features of the cell, moves the infrared lamp, adjusts the power of the infrared lamp, and drives the welding head to perform welding operations on the edge of the cell; To realize a cell welding method based on infrared lamp power control as described in any one of claims 1-7.

Images