Visual detection method for air tightness of single battery assembly of flow battery

Abstract

The invention provides an air tightness visual detection method for a single battery assembly of a flow battery. The air tightness visual detection method comprises the following steps: installing the flow battery in a detection cavity, filling a detection medium into the flow battery, and maintaining the pressure; collecting detection medium information in the flow battery; processing the acquired detection medium information through an upper computer, integrating numerical values, and then establishing comparison coordinates; and judging whether the leakage point is a leakage point or not by comparing the coordinates, and outputting a visual report. According to the gas tightness visual detection method for the single battery assembly of the flow battery, tracer gas is introduced into a closed environment, and visual detection and data processing means are utilized, so that visualization of a leakage track and automatic positioning of a leakage point are realized, the limitation that a traditional pressure maintaining / pressure drop method can only judge whether leakage exists or not is broken through, and the detection accuracy is improved. Visualization and quantitative positioning of a leakage track are realized.

Description

Technical Field

[0001] This invention belongs to the field of battery state detection technology, and in particular relates to a visual detection method for the airtightness of a single-cell flow battery assembly. Background Technology

[0002] In the production of flow battery stacks, laser welding is often used to form integrated single-cell modules. However, micro-welding defects are prone to occur during laser welding, leading to potential internal leakage risks. Therefore, it is necessary to perform airtightness testing on the laser-welded single-cell modules. Conventional airtightness testing methods often employ the pressure holding method and the differential pressure method, such as CN222070042U and 202411607595.7. The technical principle is to fill the single-cell module with gas at a certain pressure, and then seal the gas outlet through a valve or seal to form a closed system, maintaining the gas in a static state for a set time. During the test, pressure gauges or pressure sensors are used to continuously record changes in the internal gas pressure: if the pressure remains constant within the specified time, it is determined that there is no leakage; if the pressure drops, it indicates that there is a gas leakage in the module. This method typically includes a gas source device, pressure regulating valve, connecting pipelines, pressure gauges or sensors, etc., and completes the test of its sealing performance by filling, sealing and monitoring the pressure of the module under constant temperature or ambient temperature conditions. By filling a battery with a certain amount of gas and then observing the pressure change inside the battery using a pressure gauge, it is possible to determine whether there is a leak and the amount of leakage.

[0003] like Figure 2 As shown, the internal structure of a single-cell module is a stacked structure consisting of an electrode frame 100, a separator 200, a bipolar plate 300, and a carbon felt 400. The gas diffusion path within this structure is complex. Conventional pressure holding or pressure drop methods are insufficient to reveal the correspondence between the actual leakage path and weld defects. They can only determine the presence of a leak by monitoring pressure changes, failing to visually display the gas flow trajectory within the module or pinpoint the exact location of the leak. This is because such methods rely solely on changes in overall gas pressure as the detection basis, lacking the ability to visually monitor the gas diffusion path within a single cell. This inability to locate the leak point makes subsequent module repair and process improvements difficult to implement on a specific basis, ultimately leading to the module being scrapped.

[0004] Furthermore, the flow battery single cell assembly contains flexible components such as a separator 200, bipolar plates 300, and carbon felt 400. When performing internal leakage detection, gas is usually introduced into one side of the battery to pressurize it, while the other side is not pressurized. At this time, even if there is no leakage point on the pressurized side, the deformation and expansion of the flexible components inside the battery under the action of gas pressure will still cause the pressure on the pressurized side to drop, so it is impossible to determine whether the reason for the pressure drop is a leak. Summary of the Invention

[0005] Therefore, the present application aims to provide a kind of liquid flow battery cell assembly air-tightness visual detection method to solve the problem that the pressure of pressurized side is reduced due to the deformation and expansion of flexible components in the battery under the action of gas pressure when the gas is introduced into one side of the battery and the other side is not pressurized in the traditional detection method.

[0006] To achieve the above purpose, the technical scheme of the present application is as follows: A kind of liquid flow battery cell assembly air-tightness visual detection method, comprising the following steps: step one: install liquid flow battery in detection cavity, and fill detection medium into liquid flow battery and keep pressure;Step two: collect the detection medium information in liquid flow battery;Step three: the detection medium information collected is processed by host computer, and comparison coordinates are established after integration;Step four: whether it is a leakage point is judged by comparison coordinates, and output visual report.

[0007] Further, the detection medium is helium, air or tracer gas with infrared response characteristics.

[0008] Further, the method of collecting the detection medium information in the liquid flow battery in step two is to scan the welds section by section by helium mass spectrum probe or thermal imaging probe, and collect the concentration signal or infrared image signal of the detection medium in real time.

[0009] Further, the sampling frequency of the helium mass spectrum probe is 20-5000Hz, and the frame rate of the thermal imaging probe is 15-120fps;Preferably, the sampling frequency of the helium mass spectrum probe is 100Hz, and the frame rate of the thermal imaging probe is 30fps.

[0010] Further, the method of establishing comparison coordinates after integrating values in step three is to transmit the collected signal to the host computer, perform baseline establishment and background subtraction, and then execute median filtering, Gaussian filtering and contrast enhancement processing on the signal, extract the plume motion direction and speed by optical flow analysis, and obtain the center position, diffusion radius and peak intensity of the plume by two-dimensional Gaussian fitting modeling;The plume center coordinates are spatially corresponding to the pre-calibrated weld path, the minimum distance d of the plume center and the weld curve is calculated, and the leakage point is judged by the value of d.

[0011] Further, the filtering method is any one of median filtering, Gaussian filtering and bilateral filtering;Or joint processing of any two of median filtering, Gaussian filtering and bilateral filtering;Preferably, the filtering method is joint processing of median filtering and Gaussian filtering.

[0012] Further, the two-dimensional Gaussian fitting algorithm adopts the Levenberg-Marquardt nonlinear optimization method, the initial long axis radius is 8-20 pixels, the short axis radius is 3-8 pixels, the upper limit of iteration is 50 times, and convergence is achieved when the residual change is less than 1%.

[0013] Further, the calculation of the minimum distance d is based on the Euclidean distance between the plume center point and the weld path: wherein L is the total arc length of the weld path, (x(s), y(s)) is the weld path coordinates, is the plume center coordinates.

[0014] Further, when the spacing of multiple leakage points on the weld arc length coordinates is less than or equal to a set threshold, the system aggregates them into a continuous leakage interval and outputs the interval start and end coordinates.

[0015] Further, the leakage intensity classification is based on the plume peak signal intensity, diffusion radius and curve amplitude of the detection medium, and is divided into three levels of slight, moderate and severe; the plume trajectory, leakage point and leakage interval are displayed on the weld path with visual markers.

[0016] Compared with the prior art, the gas tightness visual detection method of the single cell assembly of the flow battery has the following beneficial effects: (1) The gas tightness visual detection method of the single cell assembly of the flow battery fills high-purity helium in the flow battery and forms a stable diffusion field in a controlled cavity environment, so that any small leakage in the flow battery can produce a gas plume in the weld area which can be captured.

[0017] By arranging a movable helium mass spectrometer probe or a thermal imaging probe along the weld path outside the flow battery, combined with a linear displacement structure to realize step-by-step scanning, the concentration change or temperature anomaly signal of the leakage area can be collected in real time, so as to convert the invisible local defect into a visual plume trajectory. This design ensures that the detection process has stable gas driving conditions and high spatial and temporal resolution signal input, providing a reliable physical basis for subsequent accurate identification.

[0018] (2) The gas tightness visual detection method of the flow battery single cell assembly provided by the application establishes a technical link from the original signal to the leakage positioning through a multi-step data processing method. First, the method of baseline establishment and background subtraction is used to eliminate environmental noise interference, so that the detection result only reflects the real leakage effect; second, median filtering, Gaussian filtering or bilateral filtering algorithm is used to enhance and denoise the signal, and the histogram equalization is combined to improve the contrast of the plume area and the background; then, the motion direction and speed of the plume in the continuous image are extracted through the optical flow analysis, and the stable plume and the transient false signal are distinguished; finally, the two-dimensional Gaussian modeling is used to extract the plume center, the main axis direction and the diffusion radius and other parameters, and the complex image information is converted into quantifiable mathematical representation. Through the organic combination of these data processing steps, the plume signal can be accurately extracted and parameterized, which lays a solid foundation for spatial positioning.

[0019] (3) The gas tightness visual detection method of the flow battery single cell assembly provided by the application establishes the arc length parameterized curve of the weld seam path, and maps the plume center point and the weld seam path in space, and realizes the automatic judgment of the leakage point by using the minimum distance criterion. When the minimum distance between the plume center and the weld seam path is less than or equal to 0.2-0.5mm, it is judged that leakage occurs at this place; when the distance is greater than 0.5mm and less than or equal to 2.0mm, it is marked as a suspected leakage point and triggers a secondary scan; if the distance is greater than 2.0mm, it is judged as a non-weld seam related leakage. This principle not only realizes the millimeter level leakage point positioning, but also can identify the continuous leakage interval through clustering analysis, and output the leakage coordinate and intensity grading information.

[0020] (4) The gas tightness visual detection method of the flow battery single cell assembly provided by the application solves the technical problem that the prior art can only qualitatively judge "whether there is leakage" and cannot realize quantitative positioning through the synergistic effect of the construction of the gas diffusion field, the step-by-step scanning collection of the probe and the multi-step data processing and modeling, and achieves the technical effects of leakage trajectory visualization, high-precision positioning of the leakage point and leakage intensity grading. This method realizes the transformation of the flow battery gas tightness detection from qualitative to quantitative and from invisible to visualization, and has outstanding innovation and engineering application value.

[0021] (5) The air tightness visual detection method of the flow battery single cell assembly provided by the application realizes the transformation from the traditional macroscopic judgment of "whether to leak" to the refined detection mode of "leakage trajectory visualization and automatic positioning of leakage points" through the overall concept of multi-point filling of detection medium-detection probe scanning-image processing-target detection. The method significantly improves the identification sensitivity of micro leakage and slow leakage, solves the technical problem that it is difficult to locate the leakage point in the flow battery single cell assembly due to the complex welding seam and hidden defects. At the same time, the quantitative detection result provided by the application provides a reliable technical basis for laser welding process optimization, defect repair and product quality control. BRIEF DESCRIPTION OF DRAWINGS

[0022] The drawings forming part of the present application serve to provide further understanding of the present application, and the illustrative embodiments of the present application and the description thereof serve to explain the present application and do not constitute undue limitation on the present application. In the drawings: Figure 1 The flow chart of the air tightness visual detection method of the flow battery single cell assembly according to the embodiment of the present application; Figure 2 The schematic diagram of the explosion structure of the flow battery single cell assembly according to the embodiment of the present application; Figure 3 The overall structure schematic diagram of the air tightness visual detection device of the flow battery single cell assembly according to the embodiment of the present application; Figure 4 The structure schematic diagram of the detection assembly according to the embodiment of the present application; Figure 5 The structure schematic diagram of the detection bin according to the embodiment of the present application; Figure 6 The structure schematic diagram of the mounting unit according to the embodiment of the present application; Figure 7 The structure schematic diagram of the fixing unit according to the embodiment of the present application; Figure 8 The assembly structure schematic diagram of the detection bin and the detection unit according to the embodiment of the present application; Figure 9 The structure schematic diagram of the input assembly according to the embodiment of the present application; Figure 10 The detection result schematic diagram of the flow battery by the helium source and the helium mass spectrum scanning probe according to the embodiment 1 of the present application; Figure 11 The detection result schematic diagram of the flow battery by the high-temperature air source and the thermal imaging scanning probe according to the embodiment 2 of the present application; Figure 12The detection result 1 of the flow battery by the embodiment 3 of the present application is shown in the schematic diagram of a helium source and a helium mass spectrum scanning probe; Figure 13 The detection result of the flow battery by the embodiment 4 of the present application is shown in the schematic diagram.

[0023] Explanation of reference signs: 1-detection assembly; 11-detection bin; 111-groove; 112-first sliding groove; 12-mounting unit; 121-bottom plate; 122-positioning column; 123-mounting groove; 13-fixing unit; 131-mounting plate; 132-telescopic air cylinder; 133-pressing plate; 14-detection unit; 141-cross beam; 142-first linear displacement structure; 143-second linear displacement structure; 144-first fixing block; 145-detection probe; 2-input assembly; 21-gas cylinder; 22-pressure gauge; 23-regulating valve; 24-first air pipe; 25-second air pipe; 26-mass flow meter; 3-data processing unit; 100-electrode frame; 200-separator; 300-bipolar plate; 400-carbon felt. DETAILED DESCRIPTION

[0024] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0025] In the description of the present application, it should be understood that the terms "center", "longitudinal", "transverse", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application. In addition, the terms "first", "second" and the like are only for the purpose of description and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with "first", "second" and the like can explicitly or implicitly include one or more of the features. In the description of the present application, unless otherwise stated, the meaning of "a plurality of" is two or more.

[0026] In the description of the present application, it should be noted that unless otherwise explicitly specified and limited, the terms "mounting", "connection", "linking" should be understood in a broad sense, for example, it can be fixed connection, or detachable connection, or integral connection; it can be mechanical connection, or electrical connection; it can be direct connection, or indirect connection through intermediate medium, or internal communication of two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0027] The present application will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0028] The present application provides a kind of liquid flow battery cell subassembly airtightness visual detection method, it is suitable for the liquid flow battery cell subassembly formed by laser welding sealing.

[0029] As Figure 2 Shown, liquid flow battery is generally formed by welding electrode frame 100, diaphragm 200, bipolar plate 300 and carbon felt 400 laminated structure, its weld path is long and space is complex, there is the risk of micro leakage. The method of the present application realizes the visualization of leakage track and the automatic positioning of leakage point by introducing tracer gas in a closed environment and using visual detection and data processing means.

[0030] As Figure 3As shown in the figure, a kind of liquid flow battery single cell assembly air-tightness visual detection device, including detection component 1, input component 2 and data processing unit 3, one side of detection component 1 is provided with input component 2, and one end of input component 2 is located in detection component 1, and input component 2 is used to transmit detection medium, the other side of detection component 1 is provided with data processing unit 3, and data processing unit 3 is electrically connected to detection component 1, by setting the data processing unit 3 of electric signal connection to detection component 1, the information of detection medium detected by detection component 1 can be transmitted to data processing unit 3, which is processed by data processing unit 3, and the detection result is directly displayed, data processing unit 3 is used to carry out background deduction, filter enhancement, light flow analysis and Gaussian modeling to probe signal, and the positioning and result output of leakage point are realized in combination with weld path mapping, the data processing unit 3 is modularly composed of analog-digital conversion unit, graphics processing unit, data processing unit, PLC control unit and visual interactive interface in the prior art, the analog-digital conversion unit adopts high-speed analog-digital converter with sampling frequency of 1000 Hz and resolution not less than 16 bit;The graphics processing unit is configured with embedded graphics processor, and supports ≥1 TFLOPS image operation performance;The data processing unit adopts ARM Cortex-A72 or equivalent four-core processor, and the main frequency is not less than 1.8GHz, is equipped with 8 GB memory and 64 GB solid state storage module;PLC control unit is high-speed logic programmable controller, communication interface is RS-485 / Modbus RTU, I / O point number is not less than 32 ways, and is used to cooperatively control scanning probe, gas source module and slide rail motor;Visual interactive interface is 10.1 inch touch industrial screen, and the resolution is not less than 1280×800.

[0031] As Figure 4 As shown in the figure, detection component 1 includes detection bin 11, mounting unit 12, fixed unit 13 and detection unit 14, mounting unit 12 is arranged in detection bin 11, and one side of mounting unit 12 is connected to one end of input component 2, fixed unit 13 is arranged in detection bin 11, fixed unit 13 is located above mounting unit 12, and one end of fixed unit 13 can be close to or away from one end of mounting unit 12, detection unit 14 is slidably arranged in detection bin 11, and detection unit 14 is located above fixed unit 13, the material of the detection bin 11 in the embodiment is aluminum alloy material in the prior art, and copper plate, titanium alloy or stainless steel and other metal materials with good thermal conductivity and rigidity can also be selected.

[0032] As Figure 5As shown, the detection bin 11 is a mouth-shaped structure, and the inner walls of the two sides of the detection bin 11 are respectively provided with a groove 111, and the inner walls of one side of each groove 111 are respectively fixedly connected to the two ends of the fixed unit 13. The inner walls of the two sides of the detection bin 11 are respectively provided with a first sliding groove 112, and the first sliding groove 112 is located above the groove 111, and the detection unit 14 is arranged between the two first sliding grooves 112. By arranging the groove 111 on the inner wall of the detection bin 11, the installation position of the fixed unit is limited, and the relative position of the fixed unit to the battery to be detected is ensured. By arranging the first sliding groove 112, the detection unit 14 installed in the first sliding groove 112 is arranged in sliding mode, so that the detection unit can move in the detection bin 11 for detection.

[0033] As shown in the figure, Figure 6 The installation unit 12 includes a bottom plate 121 and a positioning column 122. One side of the bottom plate 121 is fixedly connected to one end of the input assembly 2. The upper end of the bottom plate 121 is provided with a mounting groove 123, and the battery can be detachably installed in the mounting groove 123. Four positioning columns 122 are arranged in the circumferential direction in the mounting groove 123, and the periphery of each positioning column 123 is located on one side of the inner ring of the battery. The positioning column 122 is used to limit the relative position of the battery in the mounting groove 123. The bottom plate 121 has a certain thickness, and the thickness is not less than the thickness of the battery to be detected. The shape of the mounting groove 123 is consistent with the shape of the battery to be detected, and the inner ring of the mounting groove is larger than the periphery of the battery to be detected by 1mm. The depth of the mounting groove is smaller than the thickness of the battery to be detected by 2-4mm. The outer diameter, number and position of the positioning column 122 are consistent with the positioning hole of the measured assembly. The material of the positioning column 122 is EPDM or nitrile rubber or fluorine rubber or other materials in the prior art, preferably EPDM. The inner side wall of the mounting groove 123 is provided with a plurality of gas supply ports in the circumferential direction, and the output end of the output unit 2 can be connected to the battery to be detected through the gas supply port of the inner wall of the mounting groove 123 and form a loop.

[0034] As shown in the figure, Figure 7As shown, the fixing unit 13 comprises mounting plates 131, telescopic cylinders 132 and pressing plates 133, the two mounting plates 131 are fixedly connected to the inner walls of one side of the groove 111 respectively, and the two mounting plates 131 are arranged in parallel with each other, the two ends of each mounting plate 131 are fixedly connected to the inner walls of the two sides of the detection bin 11 respectively, a plurality of telescopic cylinders 132 are arranged on one side of each mounting plate 131 in the axial direction, and the movable ends of each telescopic cylinder 132 are fixedly connected to one side of the pressing plate 133 respectively, the pressing plate 133 in the embodiment is located 10-13 mm above the mounting unit 12, the material of the pressing plate 133 is transparent glass in the prior art, and the thickness is 20-30 mm, the telescopic cylinder 132 is a prior art, and the parameters thereof need to meet the following conditions: cylinder diameter Φ20-25 mm, stroke range 20-30 mm, rated working pressure 0.4-0.6 MPa, and response time less than 0.2 s. The cylinder has an adjustable stroke buffer structure, the material of the end cover is aluminum alloy, and the surface of the piston rod is subjected to hard oxidation treatment.

[0035] As shown in Figure 8 As shown, the detection unit 14 comprises a cross beam 141, a first linear displacement structure 142, a second linear displacement structure 143, a first fixed block 144 and a detection probe 145, the two ends of the cross beam 141 are peripherally and slidably connected to the inner ring of one first sliding groove 112, the first linear displacement structure 142 is arranged in any one first sliding groove 112, and the first linear displacement structure 142 is used to drive the cross beam 141 to axially displace along the first sliding groove 112, the second linear displacement structure 143 and the first fixed block 144 are arranged on the cross beam 141 respectively, and the second linear displacement structure 143 is used to drive the first fixed block 144 to axially displace along the cross beam 141, the detection probe 145 is arranged at the lower end of the first fixed block 144, and the detection probe 145 is located above the pressing plate 133, the first linear displacement structure 142 and the second linear displacement structure 143 can drive the detection probe 145 mounted on the first fixed block 144 to realize planar motion, and the detection probe 145 is located 40-45 mm above the pressing plate 133, the detection probe 145 can detect the battery to be detected through the pressing plate 133 of transparent glass material, the first linear displacement structure 142 and the second linear displacement structure 143 are both motor and screw combination in the prior art, and other linear displacement modules capable of realizing linear displacement action can also be adopted, as long as the xy direction motion of the detection probe 145 can be realized, the detection probe 145 is connected to the data processing unit 3 through an electrical signal, and the data processing unit 3 can control or observe the detection mode or detection result of the detection probe 145, the detection probe 145 can select a helium mass spectrum scanning probe or a thermal imaging scanning probe in the prior art according to the actual situation. The helium mass spectrum scanning probe uses a portable probe with air extraction, and the parameters meet the following conditions: gas pressure range 10 -1 ~103 Pa, the probe caliber is Φ3-Φ5mm, the suction flow range is 0.5-1.0L / min, and the output signal is 0-10V analog signal or RS-485 digital signal. The thermal imaging scanning probe uses a microbolometer array type probe, and the parameters meet that the detection waveband is 7.5-14μm, the resolution is not less than 640*480 pixels, the frame frequency is 30-60Hz; the probe field of view angle is 25*20°, the temperature measurement range is-20-80℃, the response time is less than 50ms, and the output interface is USB3.0.

[0036] As shown in Figure 9 , the input assembly 2 includes a gas cylinder 21, a pressure gauge 22, an adjusting valve 23, a first gas pipe 24, a second gas pipe 25 and a mass flowmeter 26, the gas cylinder 21 is arranged at one side of the detection bin 11, and the output end of the gas cylinder 21 is provided with the pressure gauge 22, the output end of the pressure gauge 22 is fixedly connected to one end of the first gas pipe 24, the middle part of the first gas pipe 24 is provided with the mass flowmeter 26, and the other end of the first gas pipe 24 is connected to the middle part of one side of the second gas pipe 25, both ends of the second gas pipe 25 are fixedly connected to one side of the bottom plate 121, and two adjusting valves 23 are arranged in the second gas pipe 25, and the two adjusting valves 23 are respectively located at the two sides of the first gas pipe 24. By arranging the pressure gauge 22, the adjusting valve 23 and the mass flowmeter 26, the input gas medium is controlled to prevent the gas medium from being unevenly distributed after entering the battery to be detected, thereby affecting the detection result. The pressure gauge 22, the adjusting valve 23 and the mass flowmeter 26 are respectively prior art. The pressure gauge 22 can be a WIKAP GS21 series mechanical pressure gauge, and the parameters need to meet that the range is 0-0.6MPa and the accuracy level is 1.6. The adjusting valve 23 can be an SMC ITV2000 series electronic proportional pressure regulating valve, and the parameters need to meet that the control accuracy is not less than ±0.5%. The mass flowmeter 26 can be a Bronkhorst EL-FLOW Select series thermal mass flowmeter, and the parameters need to meet that the range is 0-5SLM and the accuracy is ±1%F.S.

[0037] As shown in Figure 1 , a flow battery single cell assembly airtightness visual detection method flow chart: , the method comprises the following steps: Step one: install the flow battery single cell assembly sealed by laser welding in the installation unit 12, fill high-purity helium into the flow battery single cell assembly through the input assembly 2, maintain stable gas supply under 0.1-0.2MPa pressure, and make the gas form a diffusible field in the laminated structure and the weld area of the flow battery single cell assembly. Step two: start the detection probe 145, which is a helium mass spectrum probe or a thermal imaging probe, installed 5 mm above the weld path and scanned along the weld path at a speed of 2 mm / s, collecting helium concentration signals or infrared image signals in real time; Step three: transmit the collected signals to the data processing unit 3, first establish the baseline and subtract the background, then perform median filtering, Gaussian filtering and contrast enhancement processing on the signals, then extract the plume motion direction and speed through optical flow analysis, and obtain the center position, diffusion radius and peak intensity of the plume by two-dimensional Gaussian fitting modeling; Step four: spatially correspond the plume center coordinates with the pre-calibrated weld path, calculate the minimum distance d between the plume center and the weld curve, when d≤0.5 mm, determine that the position is a leakage point, when 0.5 mm<d≤2.0 mm, mark it as a suspected leakage point, when d>2.0 mm, determine it as a non-weld leakage; Step five: output the visual report containing the coordinate position of the leakage point or leakage interval, plume image and leakage intensity classification on the detection result interface.

[0038] Data acquisition and processing The detection process was carried out in an environment with temperature of 25±1℃ and humidity control of 40%–60%. The data acquisition and processing process of the air tightness detection was carried out according to the following steps: Background subtraction Firstly, the flow battery cell assembly is placed in the installation unit 12, and positioned by the positioning hole of the flow battery cell assembly and the positioning column 122. Then the pressing plate 133 is pressed down under the action of the telescopic cylinder 132 to press the flow battery cell assembly. Before filling the detection medium, the data processing unit 3 runs the detection probe 145 for 30-60 seconds to establish a baseline. The detection probe 145 outputs the helium concentration or thermal imaging curve at a sampling rate of 1000 Hz, and a stable baseline is calculated by using a sliding average (window size 5-10 s) or an exponential sliding average (decay coefficient 0.01-0.05). Then, during the detection process, the real-time concentration signal is subtracted from the baseline value point by point to ensure that the final input data only reflect the true signal caused by leakage, avoiding false judgments caused by environmental noise. Then, through the input assembly 2, the detection medium is filled, and the pressure is set to 0.1-0.2 MPa during the process of filling the detection medium. The gas temperature and type are determined according to the type of the detection probe 145. The second gas pipe 25 and the inner side wall of the installation groove 123 are provided with multiple gas supply ports in the circumferential direction to simultaneously supply gas to the positive and negative electrode flow channel areas of the flow battery cell assembly. After 2 minutes of balance, the gas is fully diffused in the laminated structure and the weld area. Then, if the detection probe 145 is a thermal imaging probe, the first linear displacement structure 142 and the second linear displacement structure 143 jointly drive the detection probe 145 to be fixed at a position above the center of the flow battery cell assembly to take pictures, obtaining the plume image of the gas temperature distribution. If the detection probe 145 is a helium mass spectrometer probe, it scans the flow battery cell assembly globally and segmentally at a speed of 2 mm / s to form a plume image of the helium concentration distribution.

[0039] Image preprocessing The plume image collected by the detection probe 145 is input into the data processing unit 3 through the signal line. First, it is converted into a digital signal by an analog-to-digital converter and enters the image processing stage. The sampling frequency is set to 1000 Hz in the processing stage. First, it needs to be enhanced and filtered. The optional filters include a median filter (kernel size 3x3 or 5x5, used to remove isolated bright spots), a Gaussian filter (σ=0.5-2.0 pixels, used to smooth noise), and a bilateral filter (spatial kernel size 5-9 pixels, intensity difference threshold 15-50, used to remove noise while preserving the plume edge), preferably a combination of median and Gaussian processing to balance edge preservation and noise suppression. Then the image is histogram equalized or CLAHE (contrast limited adaptive histogram equalization, block size 8x8, clipping limit 2.0) to improve the contrast between the plume area and the background. In the time domain, frame averaging is used with a window length of 3-5 frames, corresponding to about 50-100 ms, to reduce high-frequency flicker noise. After processing, the plume boundary is sharper and the gray scale distribution is smoother, which is beneficial for subsequent motion feature extraction.

[0040] Motion feature extraction The displacement vector field of the plume pixel region is extracted by comparing two consecutive images, and the average motion direction is calculated. The sampling frame rate is set to 30-60 fps, the time interval is 16-33 ms, and the vector field resolution is set to 8x8 or 16x16 pixel blocks. If a certain region maintains the same motion direction for 5-10 consecutive frames (about 0.1-0.3 s) and the speed amplitude is between 0.5-3.0 pixels / frame, it is determined to be a stable plume signal. Otherwise, if the signal lasts only for 1-2 frames and the direction is irregular, it is considered to be a transient noise and is automatically removed. Through this method, more than 90% of random false signals can be effectively filtered out, and the plume displacement vector field is obtained, which provides a reliable motion principal axis direction for plume modeling.

[0041] Plume modeling After determining the stable existence of the plume, the data processing unit 3 models its morphology to generate a parameterized description. First, a window of size 64x64 or 128x128 pixels is cut in the region where the plume is located, and the window size is flexibly adjusted according to the camera resolution, usually covering a physical range of 1-5 mm of the weld 1. Subsequently, a fitting algorithm is used to approximate the plume shape to a two-dimensional Gaussian distribution, outputting the center coordinates, principal axis direction, diffusion radius and peak intensity of the plume. The fitting algorithm can select least squares, Levenberg-Marquardt nonlinear optimization or expectation maximization Gaussian mixture modeling, and the Levenberg-Marquardt algorithm is preferred. The initial value of the fitting is set to a major axis radius of 8-20 pixels and a minor axis radius of 3-8 pixels, and the upper limit of the number of iterations is 50 times, and the convergence condition is that the residual change is less than 1%. In actual operation, the fitting process is repeated every 0.5 seconds to ensure dynamic updating of the results. When the peak intensity of the plume is less than 1.5 times the background mean value, it is determined to be a weak plume and does not enter the next mapping. This modeling method can convert complex images into a set of stable parameters, providing accurate input for spatial positioning.

[0042] Weld mapping Weld mapping is achieved by spatially corresponding the plume center position obtained by plume modeling with the pre-designed weld path. Before detection, the weld path is established as an arc length parameterized curve by a slide rail encoder or a laser ranging device, and the calibration accuracy is preferably controlled within ±0.5 mm, more preferably ±0.2 mm. During detection, the system converts the plume center pixel coordinates into physical coordinates through camera calibration parameters, and calculates the minimum distance d to the weld curve, as shown in equation (1).

[0043]

[0044] In the formula, d represents the minimum distance between the plume center point and the weld path, and the symbol (x(s), y(s)) represents the spatial position coordinates of the weld path; the symbol A spatial position coordinate representing a plume center.

[0045] Leak point determination When the distance d is less than or equal to a preset threshold of 0.2-0.5 mm, it is determined that the weld position corresponding to the plume point has a leak, and its coordinate position on the weld arc length is recorded; when the distance is greater than 0.5 mm and less than or equal to 2.0 mm, the point is marked as a suspected weld-related leak, and a secondary fine detection can be triggered, preferably repeated scanning within a range of ±5-10 mm with a step distance of 0.5-1.0 mm; when the distance is greater than 2.0 mm, it is determined that the plume source does not belong to the weld area, but to the internal cavity of the flow battery cell assembly. Further, a plurality of mapping points can be clustered according to the arc length coordinates, and when the spacing between several points is not greater than 5-10 mm, they are combined into a continuous leak interval, and the start and end arc length coordinates of the interval are output. Thus, after the weld mapping step, the precise positioning of the leak point or leak interval on the weld path can be achieved, and the positioning accuracy is limited to a range of ±0.2-0.5 mm. Finally, the helium concentration or thermal imaging image after image preprocessing is output on the data processing unit 3, the leak point and leak interval are marked on the image, and the leak point and leak interval coordinates are displayed.

[0046] The present application provides four embodiments.

[0047] As Figure 10 shown, embodiment 1 provides a flow battery cell assembly airtightness visualization detection method based on helium tracer and helium mass spectrometry leak detection; In this embodiment, the size of the measured liquid flow battery cell assembly is 600 mm x 400 mm x 9 mm, and the detection process is carried out in an environment with a temperature of 25±1℃ and a humidity control of 40%-60%. The gas cylinder 21 adopts a high-purity helium cylinder, and the gas purity is ≥99.999%. The charging pressure is set to 0.1-0.15 MPa, and the flow rate is controlled to be 0.3-0.8 L / min. The helium gas enters the installation groove 123 through the pressure regulating valve 23 and the mass flow meter 26, and is injected into the detection medium in the positive and negative electrode flow channel areas of the liquid flow battery cell assembly through the gas supply interfaces distributed circumferentially in the installation groove 123. After 2 minutes of equilibrium diffusion, the gas is fully diffused inside the electrode stack and the weld. The detection probe 145 is a helium mass spectrometry probe, which is fixed in the first fixed block 144 above the cavity, with a probe distance of 5 mm from the weld surface. The device is operated for 30 seconds without gas to establish a baseline signal before detection, and the sliding average method is used to eliminate environmental noise. During the detection process, the real-time concentration signal is input into the data processing unit 3 after being deducted from the baseline. The data processing unit 3 processes the signal sequence using median filtering, Gaussian filtering, and histogram equalization, extracts the plume dynamic characteristics combined with optical flow analysis, and extracts the plume center, diffusion radius, and intensity through two-dimensional Gaussian fitting. Subsequently, the data processing unit 3 spatially corresponds the plume center to the weld path and calculates the minimum distance d, which is determined as a leakage point when d≤0.5 mm.

[0048] The detection results show that there are 2 leakage points in the range of about 1.9 m of the weld length, which are located at 860-868 mm from the start of the weld. The corresponding helium concentration signal peak is significantly higher than 3 times the baseline, and the positioning accuracy is controlled within ±1 mm.

[0049] As shown in Figure 11 , Example 2 provides a liquid flow battery cell assembly air tightness visual detection method based on high-temperature air tracing and thermal imaging leak detection.

[0050] In this embodiment, the measured liquid flow battery cell assembly is the same as in Example 1, the detection device structure is basically the same as in Example 1, and the detection environment temperature and humidity are the same as in Example 1. However, the detection probe 145 is replaced with an infrared thermal imaging probe. The detection medium uses heated high-purity air with a temperature of 40±2℃, and the charging pressure is controlled to be 0.1-0.15 MPa, and the flow rate is controlled to be 0.5-1.0 L / min. The gas enters the liquid flow battery positive and negative electrode flow channels through the gas supply ports on the side wall of the installation groove 123, and forms a stable gas distribution after 2 minutes of equilibrium.

[0051] The infrared thermal imaging probe is installed 40 mm above the flow battery, fixed on the first fixed block 144, and collects image sequences in a static global imaging mode. The camera resolution is 1280x720, the frame rate is 30 fps, and the exposure time is set to 10 ms. Before detection, the probe collects background images for 30 s to establish a baseline template. During the detection process, the data processing unit 3 first performs median filtering (3x3), Gaussian filtering (σ=1.0), and histogram equalization processing on the collected images, and then performs CLAHE (block size 8x8, limit value 2.0) to enhance the contrast between the plume and the background. In the time domain, 3-5 consecutive frames are averaged to reduce high-frequency flicker. Then, the dynamic direction feature of the plume is extracted by using the optical flow analysis method, and the Levenberg-Marquardt algorithm is used for two-dimensional Gaussian fitting to obtain the plume center position and diffusion radius.

[0052] In the weld mapping stage, the minimum distance d between the plume center coordinates and the weld path is calculated as a criterion, and when d≤0.5 mm, it is determined as a leakage point. The detection results show that one leakage point is identified in the edge area of the flow battery, which is located at 860-900 mm from the start of the weld, consistent with the detection results of Example 1. The plume temperature signal at this position in the thermal imaging image is 1.7 times higher than the background, and the positioning accuracy is ±1 mm.

[0053] As shown in Figure 12 , in Example 3, the size of the measured flow battery single cell assembly is 600 mm x 400 mm x 9 mm, and the flow battery is sealed by laser welding of the electrode frame 100, the diaphragm 200, the bipolar plate 300, and the carbon felt 400, but the welding position is different from the assemblies of Example 1 and Example 2. The detection device, detection process, and detection conditions are the same as those of Example 1. The detection results show that a total of two leakage points are detected within the full length of about 2.2 m of the weld, which are located at 52-55 mm and 361-374 mm from the start of the weld, respectively, and the corresponding helium concentration signal peak is significantly higher than 3 times the baseline, with a positioning accuracy of ±1 mm.

[0054] As shown in Figure 13 , in Example 4, the measured flow battery single cell assembly is the same as that of Example 3, and the detection device, detection process, and detection conditions are the same as those of Example 2. The detection results show that a total of two leakage points are detected within the full length of about 2.2 m of the weld, which are located at 52-55 mm and 361-374 mm from the start of the weld, respectively, and the corresponding plume temperature signal is higher than 2 times the background, with a positioning accuracy of ±1 mm.

[0055] The working process of a liquid flow battery single cell assembly air tightness visual detection device: Firstly, the measured liquid flow battery cell assembly is placed in the mounting groove 123 in the mounting unit 12, and is positioned with the positioning hole and the positioning column 122. Then the pressing plate 133 is pressed down under the action of the plurality of telescopic cylinders 132, and then the liquid flow battery cell assembly is filled with detection medium by the input assembly 2. The gas temperature and type are determined according to the type of the detection probe 145, and the detection medium is filled at a pressure of 0.1-0.2 MPa. The detection medium flows and diffuses to the positive and negative electrode flow channel areas of the battery assembly through the plurality of gas supply openings provided on the inner side wall of the first gas pipe 24, the second gas pipe 25 and the mounting groove 123. Then, the detection probe 145 is driven to take a picture at a position above the center of the battery assembly by the first linear displacement structure 142 and the second linear displacement structure 143, so as to obtain the information of the filled detection medium. The information of the filled detection medium collected by the detection probe 145 is input into the data processing unit 3 through the electrical signal connection, and the information detected by the detection probe 145 is processed and visually output by the data processing unit 3.

[0056] The control mode of the embodiment is controlled by a controller. The control circuit of the controller can be realized by simple programming by those skilled in the art, and the power supply also belongs to the common knowledge in the art. The present document mainly protects the mechanical device, and the control mode and circuit connection are not explained in detail.

[0057] The above only describes the preferred embodiments of the present application and is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for visually detecting the airtightness of a single-cell flow battery assembly, characterized in that: Includes the following steps: Step 1: Install the flow battery into the detection chamber, fill the flow battery with the detection medium, and maintain the pressure. Step 2: Collect information about the detection medium inside the flow battery; Step 3: The collected detection medium information is processed by the host computer, and the values ​​are integrated to establish comparison coordinates; Step 4: Determine if it is a leak point by comparing coordinates, and output a visual report.

2. The method for visually detecting the airtightness of a single-cell flow battery assembly according to claim 1, characterized in that: The detection medium is helium, air, or a tracer gas with infrared response characteristics.

3. The method for visually detecting the airtightness of a single-cell flow battery assembly according to claim 1, characterized in that: Step two involves collecting information about the detection medium within the flow battery by scanning the weld segment by segment using a helium mass spectrometer or thermal imaging probe and acquiring the concentration signal or infrared image signal of the detection medium in real time.

4. The method for visually detecting the airtightness of a single-cell flow battery assembly according to claim 1, characterized in that: The sampling frequency of the helium mass spectrometer probe is 20–5000 Hz, and the frame rate of the thermal imaging probe is 15–120 fps. Preferably, the sampling frequency of the helium mass spectrometer probe is 100 Hz, and the frame rate of the thermal imaging probe is 30 fps.

5. The method for visually detecting the airtightness of a single-cell flow battery assembly according to claim 1, characterized in that: Step 3, after integrating the numerical values, establishes the comparison coordinates by transmitting the acquired signal to the host computer, performing baseline establishment and background subtraction, and then performing median filtering, Gaussian filtering and contrast enhancement processing on the signal. The direction and velocity of the plume movement are extracted through optical flow analysis, and the center position, diffusion radius and peak intensity of the plume are obtained by using two-dimensional Gaussian fitting modeling. The coordinates of the plume center are spatially correlated with the pre-calibrated weld path, and the minimum distance d between the plume center and the weld curve is calculated. The leakage point is determined by the value of d.

6. The method for visually detecting the airtightness of a single-cell flow battery assembly according to claim 1, characterized in that: The filtering method is any one of median filtering, Gaussian filtering, and bilateral filtering; Or a combination of any two of median filtering, Gaussian filtering, and bilateral filtering; Preferably, the filtering method is a combination of median filtering and Gaussian filtering.

7. The method for visually detecting the airtightness of a single-cell flow battery assembly according to claim 5, characterized in that: The two-dimensional Gaussian fitting algorithm adopts the Levenberg-Marquardt nonlinear optimization method, with an initial major axis radius of 8-20 pixels and a minor axis radius of 3-8 pixels. The upper limit of the iteration is 50 times, and it converges when the residual change is less than 1%.

8. The method for visually detecting the airtightness of a single-cell flow battery assembly according to claim 5, characterized in that: The minimum distance d is calculated based on the Euclidean distance between the plume center point and the weld path: ； Where (x(s),y(s)) are the weld path coordinates, The coordinates are the center coordinates of the plume.

9. The method for visually detecting the airtightness of a single-cell flow battery assembly according to claim 5, characterized in that: When the distance between multiple leakage points on the weld arc length coordinate is less than or equal to a set threshold, the system aggregates them into a continuous leakage interval and outputs the start and end coordinates of the interval.

10. The method for visually detecting the airtightness of a single-cell flow battery assembly according to claim 1, characterized in that: The leakage intensity classification is based on a comprehensive assessment of the plume peak signal intensity, diffusion radius, and curve amplitude of the detection medium, and is divided into three levels: slight, moderate, and severe. The plume trajectory, leak point, and leak area are displayed as visual markers on the weld path.

Images