Pipeline flange liquid leakage detection method and system based on hyperspectral recognition

Abstract

The present application relates to pipeline leakage detection technical field, specifically to a kind of pipeline flange liquid leakage detection method and system based on hyperspectral identification, and one kind of pipeline flange liquid leakage detection method based on hyperspectral identification includes the following steps: S1: pipeline flange liquid leakage detection system initialization, emit monitoring light beam to establish the monitoring light path for monitoring whether pipeline flange has liquid leakage, when pipeline flange occurs liquid leakage and causes monitoring light path to be cut off, turn into step S2;S2: preliminary liquid leakage positioning is carried out, and the target area of suspected existence liquid leakage is positioned;S3: target area is carried out hyperspectral collection, obtains the spectral data of target area, according to the spectral data and the comparison analysis of pre-set standard liquid spectral database, the relevant information of leakage liquid is output simultaneously.The present application can realize the full-time monitoring of pipeline flange, realizes real-time accurate detection and liquid component intelligent discrimination under complex industrial environment.

Description

Technical Field

[0001] The present invention relates to the technical field of pipeline leakage detection, and in particular to a pipeline flange liquid leakage detection method and system based on hyperspectral recognition. Background Art

[0002] Pipelines play a critical role in material transportation in industrial production, particularly in industries like chemicals and energy, where thousands of pipelines crisscross and form a vast transportation network. Currently, most pipeline leak detection relies on manual inspections. However, with such a large number of pipelines, manual inspections not only have a low problem detection rate but also an extremely long response time. This is especially true in hazardous chemical companies. If pipeline leaks go undetected, they can cause serious accidents such as fires, explosions, and environmental pollution, resulting in immeasurable loss of life and property.

[0003] While some companies have deployed monitoring equipment to detect leaks, most existing equipment relies on RGB image recognition, which can only detect leaks based on color differences. Industrial fluids vary widely (e.g., acids and alkalis of varying concentrations, oils, solvents, etc.), and similarly colored liquids can easily lead to misidentification (e.g., it's difficult to distinguish between light yellow motor oil and transparent hydraulic oil). Furthermore, interference from factors like ambient lighting and liquid evaporation prevents accurate identification of the composition and properties of leaking liquids. While this type of equipment can provide initial warnings for large-scale leaks, it often misses slow leaks or small leaks due to insufficient pixel resolution or blurred color features. Furthermore, relying on manual inspections of monitoring footage still prevents real-time location of leaks and risk assessment, making overall detection efficiency and accuracy insufficient to meet industry requirements. Breaking through the bottlenecks of traditional technologies to achieve real-time, accurate detection of small leaks and intelligent identification of liquid composition in complex industrial environments remains a pressing challenge. Summary of the Invention

[0004] The technical problem to be solved by the present invention is that the existing pipeline inspection work relies on manual inspection, has a low degree of automation, and relies on human judgment, has a high misjudgment rate and missed detection rate, and is not real-time, posing a major safety hazard.

[0005] To solve the above technical problems, the first aspect of the present invention adopts the following technical solution: a method for detecting liquid leakage of a pipe flange based on hyperspectral recognition, comprising the following steps:

[0006] S1: Initialize the pipeline flange liquid leakage detection system and emit a monitoring light beam to establish a monitoring light path for monitoring whether there is liquid leakage in the pipeline flange. When liquid leakage occurs in the pipeline flange and the monitoring light path is cut off, the system proceeds to step S2;

[0007] S2: Perform preliminary fluid leakage localization to locate the target area where fluid leakage is suspected;

[0008] S3: Perform hyperspectral acquisition on the target area to obtain spectral data of the target area, and compare and analyze the spectral data with a preset standard liquid spectrum database to determine whether there is liquid leakage in the target area. If there is liquid leakage, relevant information of the leaking liquid is output synchronously.

[0009] When the present invention is working, it can realize a series of tasks such as full-time monitoring of pipeline flanges, liquid leakage monitoring and early warning, liquid leakage monitoring and determination work and liquid sufficient intelligent judgment. It has a high degree of automation and intelligence, reduces the proportion of manual participation, and has high real-time and high accuracy of detection. It can effectively improve the overall detection efficiency and accuracy, and realize real-time and accurate detection of tiny leakage points and intelligent judgment of liquid composition in complex industrial environments.

[0010] Preferably, in step S1, when the pipeline flange liquid leakage detection system is initialized and a monitoring light beam is emitted to establish a monitoring light path for monitoring whether the pipeline flange has liquid leakage, the following steps are adopted:

[0011] A1: The leakage monitoring mechanism provided in the pipeline flange liquid leakage detection system is arranged around the periphery of the pipeline flange, so that the reflective portion provided in the leakage monitoring mechanism is arranged around the periphery of the pipeline flange;

[0012] A2: The leakage monitoring mechanism is equipped with an infrared generating device to emit infrared light along a preset angle. The infrared light is irradiated on the reflecting part and refracted in sequence to form an infrared light path arranged around the pipe flange. The leakage monitoring mechanism is equipped with an infrared receiving device to be arranged at the end of the infrared light path to receive the infrared light, thereby establishing a monitoring light path for monitoring whether there is liquid leakage in the pipe flange.

[0013] Preferably, in step S1, when the pipeline flange liquid leakage detection system is initialized and a monitoring light beam is emitted to establish a monitoring light path for monitoring whether the pipeline flange has liquid leakage, the following steps are adopted:

[0014] B1: Arrange the leakage monitoring mechanism provided in the pipeline flange liquid leakage detection system around the periphery of the pipeline flange;

[0015] B2: The infrared generating device and the infrared receiving device are arranged relative to each other in the leakage monitoring mechanism, and the infrared generating device reflects the infrared light to establish an infrared light path, and the infrared light path is connected to the movement path of the leaking liquid to establish a monitoring light path for monitoring whether there is liquid leakage in the pipeline flange.

[0016] Preferably, step S3 further includes the following steps: calling pre-collected pipeline model data to perform background stripping on the spectral data of the target area, and then comparing and analyzing the spectral data with a preset standard liquid spectral database to determine whether there is liquid leakage in the target area.

[0017] Preferably, the standard liquid spectrum database is established by the following steps:

[0018] D1: Obtain several types of pipeline transported liquids and collect the original data of each pipeline transported liquid;

[0019] D2: Processing the raw data of each pipeline transported liquid to obtain a standard model for each pipeline transported liquid;

[0020] D3: Obtain relevant information for several practical application scenarios, assign weights to several standard models, and store them as a standard liquid spectrum database.

[0021] To solve the above technical problems, the second aspect of the present invention adopts the following technical solution: applying a pipeline flange liquid leakage detection method based on hyperspectral recognition as described in any of the above aspects, comprising:

[0022] The main control system is used for overall control, data reception and data processing, and performs comparative analysis based on the collected spectral data to output relevant information about the leaking liquid;

[0023] Leakage monitoring mechanism, used to monitor whether there is liquid leakage in the pipeline flange and output corresponding early warning signal when liquid leakage occurs in the pipeline flange;

[0024] A hyperspectral acquisition mechanism is used to perform hyperspectral acquisition on a target area to obtain spectral data of the target area;

[0025] The acquisition drive mechanism is used to drive the displacement of the hyperspectral acquisition mechanism and adjust the acquisition angle of the hyperspectral acquisition mechanism;

[0026] The leakage monitoring mechanism, the acquisition drive mechanism and the hyperspectral acquisition mechanism are all connected to the data of the main control system. The leakage monitoring mechanism is provided with a monitoring optical path for monitoring whether there is liquid leakage in the pipeline flange. The monitoring optical path is connected to the movement path of the leaking liquid. The leakage monitoring mechanism outputs a corresponding early warning signal when the monitoring optical path is cut off. After receiving the corresponding early warning signal, the hyperspectral acquisition mechanism is driven by the acquisition drive mechanism to move to the target position. When the hyperspectral acquisition of the target area suspected of liquid leakage is completed, the hyperspectral acquisition mechanism outputs spectral data to the main control system. The main control system determines whether there is liquid leakage in the target area after completing the comparative analysis of the spectral data.

[0027] Preferably, the leakage monitoring mechanism includes a first infrared generator, a first infrared receiver, a polygonal reflector group and a mounting clamp. The mounting clamp is arranged around the periphery of the corresponding pipe flange. The inner wall surface of the upper end of the mounting clamp maintains a gap with the pipe flange for leaking liquid to flow in. The polygonal reflector group is provided with a plurality of polygonal slices. The plurality of polygonal slices are fixedly installed on the inner wall surface of the upper end of the mounting clamp in sequence according to preset angles around the pipe flange. Reflective films for reflecting infrared light are provided in the plurality of polygonal slices. The first infrared generator and the first infrared receiver are both arranged inside the upper end of the mounting clamp. The first infrared generator and the first infrared receiver are arranged opposite to each other and emit infrared light to illuminate the polygonal reflector group. The polygonal reflector group refracts the infrared light to the receiving part of the first infrared receiver to establish a monitoring optical path.

[0028] Preferably, the acquisition drive mechanism includes an annular guide rail, a first sliding mount and a first drive device. The annular guide rail is arranged around the periphery of the corresponding pipe flange. The first sliding mount is slidably limited and mounted on the annular guide rail. The output part of the first drive device is transmission-connected to the first sliding mount. The hyperspectral acquisition mechanism is mounted on the first sliding mount and moves to the target position when the first drive device drives the first sliding mount to move.

[0029] Preferably, the leakage monitoring mechanism includes a second infrared generator, a second infrared receiver and a mounting rack, the second infrared generator and the second infrared receiver are both mounted on the mounting rack and arranged oppositely below the corresponding pipe flange, and the second infrared generator emits infrared light to illuminate the receiving part of the second infrared receiver to establish a monitoring optical path.

[0030] Preferably, the acquisition drive mechanism includes a translation rail, a second sliding mount and a second drive device, the translation rail is horizontally arranged below the corresponding pipe flange, the second sliding mount is slidably limitedly mounted on the translation rail, and the hyperspectral acquisition mechanism is mounted on the second sliding mount and moves to the target position when the second drive device drives the second sliding mount to move.

[0031] The beneficial technical effects of the present invention include:

[0032] The present invention can realize a series of tasks such as full-time monitoring of pipeline flanges, liquid leakage monitoring and early warning, liquid leakage monitoring and determination, and intelligent liquid sufficiency judgment. It has a high degree of automation and intelligence, reduces the proportion of manual participation, and has high real-time and high accuracy in detection. It can effectively improve the overall detection efficiency and accuracy, and realize real-time and accurate detection of tiny leakage points and intelligent judgment of liquid composition in complex industrial environments.

[0033] Other features and advantages of the present invention will be disclosed in detail in the following specific embodiments and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The present invention will be further described below with reference to the accompanying drawings:

[0035] Figure 1 This is a workflow diagram of a pipeline flange liquid leakage detection method based on hyperspectral recognition;

[0036] Figure 2 A workflow diagram for establishing a monitoring optical path in a pipeline flange liquid leakage detection method based on hyperspectral recognition;

[0037] Figure 3 This is a structural diagram of a pipeline flange liquid leakage detection system based on hyperspectral recognition;

[0038] Figure 4 Decomposition of a pipeline flange liquid leakage detection system based on hyperspectral recognition Figure 1 ;

[0039] Figure 5 for Figure 4 A partial enlarged view of point A in the middle;

[0040] Figure 6 It is a partial enlarged view of some parts of the leakage monitoring mechanism;

[0041] Figure 7 Decomposition of a pipeline flange liquid leakage detection system based on hyperspectral recognition Figure 2 (Part of the protective cover is hidden);

[0042] Figure 8 for Figure 7 A partial enlarged view of point B in the middle. DETAILED DESCRIPTION

[0043] The following is an explanation and description of the technical solutions of the embodiments of the present invention in conjunction with the drawings of the embodiments of the present invention. However, the following embodiments are only preferred embodiments of the present invention and are not exhaustive. Based on the embodiments in the implementation manner, other embodiments obtained by those skilled in the art without creative work are all within the scope of protection of the present invention.

[0044] In the following description, terms such as "inside", "outside", "up", "down", "left", "right", etc. that indicate directions or positional relationships are only used to facilitate the description of the embodiments and simplify the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operate in a specific direction. Therefore, they should not be understood as limiting the present invention.

[0045] Example 1:

[0046] See also Figure 1 This embodiment discloses a method for detecting liquid leakage of a pipe flange based on hyperspectral recognition, comprising the following steps:

[0047] S1: Initialize the pipeline flange liquid leakage detection system and emit a monitoring light beam to establish a monitoring light path for monitoring whether there is liquid leakage in the pipeline flange. When liquid leakage occurs in the pipeline flange and the monitoring light path is cut off, the system proceeds to step S2;

[0048] S2: Perform preliminary fluid leakage localization to locate the target area where fluid leakage is suspected;

[0049] S3: Perform hyperspectral acquisition on the target area to obtain spectral data of the target area, and compare and analyze the spectral data with a preset standard liquid spectrum database to determine whether there is liquid leakage in the target area. If there is liquid leakage, relevant information of the leaking liquid is output synchronously.

[0050] When this embodiment is working, it can realize a series of tasks such as full-time monitoring of pipeline flanges, liquid leakage monitoring and early warning, liquid leakage monitoring and determination work and liquid sufficiency intelligent judgment. It has a high degree of automation and intelligence, reduces the proportion of manual participation, and has high real-time and high accuracy of detection. It can effectively improve the overall detection efficiency and accuracy, and realize real-time and accurate detection of tiny leakage points and intelligent judgment of liquid composition in complex industrial environments.

[0051] See also Figure 2 Preferably, in order to adapt to pipelines with different layout forms, a corresponding leakage monitoring mechanism 2 is required. When the pipeline is arranged vertically, in step S1, the pipeline flange liquid leakage detection system is initialized and a monitoring light beam is emitted to establish a monitoring light path for monitoring whether there is liquid leakage in the pipeline flange. The following steps are adopted:

[0052] A1: The leakage monitoring mechanism 2 provided in the pipeline flange liquid leakage detection system is arranged around the periphery of the pipeline flange, so that the reflective portion provided in the leakage monitoring mechanism 2 is arranged around the periphery of the pipeline flange;

[0053] A2: The leakage monitoring mechanism 2 is provided with an infrared generating device to emit infrared light along a preset angle. The infrared light is irradiated on the reflecting part and refracted in sequence to form an infrared light path arranged around the pipe flange. The leakage monitoring mechanism 2 is provided with an infrared receiving device to be arranged at the end of the infrared light path to receive the infrared light, so as to establish a monitoring light path for monitoring whether there is liquid leakage in the pipe flange.

[0054] When the pipeline is arranged horizontally, in step S1, the pipeline flange liquid leakage detection system is initialized and a monitoring light beam is emitted to establish a monitoring light path for monitoring whether there is liquid leakage at the pipeline flange. The following steps are adopted:

[0055] B1: Arrange the leakage monitoring mechanism 2 provided in the pipeline flange liquid leakage detection system around the periphery of the pipeline flange;

[0056] B2: The infrared generating device and the infrared receiving device are arranged relative to each other in the leakage monitoring mechanism 2, and the infrared generating device reflects the infrared light to establish an infrared light path, and the infrared light path is connected to the movement path of the leaking liquid to establish a monitoring light path for monitoring whether there is liquid leakage in the pipeline flange.

[0057] During operation, when the leaked liquid invades the monitoring light path, the monitoring light path will be immediately blocked, so that the leakage monitoring mechanism 2 can immediately output a warning signal to carry out the next step of hyperspectral acquisition. As a further improvement of this embodiment, in step S1, the following steps are also included: through the data signal processing module provided in the leakage monitoring mechanism 2, the monitoring data of the sampling circuit of the data signal processing module is obtained in real time and drawn into a first curve, and the curve fluctuation and the curve fluctuation duration of the first curve are monitored according to the preset judgment rules. When the curve fluctuation of the first curve is greater than the preset threshold, and the time of the curve fluctuation exceeds the preset threshold, the reporting mechanism is activated to output a warning signal.

[0058] Among them, the first curve represents the monitoring data change curve of a single leakage monitoring mechanism 2 within a certain period of time. When working, for example, the sampling period of the sampling circuit can be set to 100ms to collect once, and the fluctuation threshold is set to 0.75. When the curve fluctuation of the first curve is greater than 0.75 and the duration exceeds 300ms, it is considered that the corresponding pipeline flange has a leak, and the reporting mechanism is activated, an early warning signal is output, and step S2 is entered.

[0059] In the specific implementation, in order to further improve the monitoring accuracy, in step S1, the following steps are also included, which summarize the monitoring data of multiple leakage monitoring mechanisms 2 and draw them into a second curve. The leakage monitoring mechanism 2 is adaptively calibrated according to the preset monitoring period through the second curve. For example, all leakage monitoring mechanisms 2 can be periodically calibrated every ten seconds. Of course, when abnormal data fluctuations occur in a single leakage monitoring mechanism 2, calibration can be automatically performed, so that this embodiment can work in harsh environments with changing conditions, and can reduce the number of adaptive maintenance times, with strong adaptability and low maintenance costs.

[0060] Preferably, the actual service life D of the flange sealing ring can be determined by collecting the sealing ring replacement records in the maintenance record sheet. Based on this data, an inspection threshold X is defined, where X=D / 2. When the preset threshold is reached, the flange point is promoted to a key monitoring point, the ADC sampling period is increased to 50ms each time, and the fluctuation threshold is adjusted to 0.5. In other words, the accuracy of Curve 1 is improved to prevent missed detections.

[0061] Example 2:

[0062] This embodiment provides a method for detecting liquid leakage in a pipeline flange based on hyperspectral recognition. The similarities with other embodiments are not repeated here, and the differences are described in detail below.

[0063] In this embodiment, in step S3, the spectral data is compared with a preset standard liquid spectrum database to determine whether there is liquid leakage in the target area. When liquid leakage is present, the following steps are used to synchronously output relevant information of the leaked liquid:

[0064] E1: Calculate the reflectance data of the spectral data, perform first-order derivative on each pixel point of a frame of spectral data, and obtain the first-order derivative data frame. When working, the number of bands of each pixel point is 300 bands, and the band values ​​corresponding to these 300 bands are respectively used To express, the reflectivity value corresponding to each band is expressed as To express;

[0065] Its first-order derivative is , slide one value in turn to get the first-order derivative of all bands of the pixel point, that is, the value of 299 first-order derivatives , and repeatedly know that the first-order derivative data frame of all pixel points of a frame of spectral data is obtained.

[0066] E2: Compare the first-order derivative curve of each pixel point in the first-order derivative data frame with the curve recorded in the standard liquid spectrum database to determine whether the spectral data matches the curve recorded in the standard liquid spectrum database. In specific implementation, a variance value can be obtained through Euclidean calculation. When the variance value is less than the set threshold, it is considered that this curve is similar to the corresponding standard sample. When the variance value is similar to multiple curves, the curve with the smallest variance value is determined as the final result and the matching result is output.

[0067] Preferably, step S3 further includes the following steps: after calling the pre-collected pipeline model data to perform background stripping on the spectral data of the target area, the spectral data is compared and analyzed with a preset standard liquid spectral database to determine whether there is liquid leakage in the target area.

[0068] In specific implementation, the following steps are used to establish the standard liquid spectrum database:

[0069] D1: Obtain several types of pipeline transported liquids and collect the original data of each pipeline transported liquid;

[0070] D2: Processing the raw data of each pipeline transported liquid to obtain a standard model for each pipeline transported liquid;

[0071] D3: Obtain relevant information for several practical application scenarios, assign weights to several standard models, and store them as a standard liquid spectrum database.

[0072] In specific implementation, the following steps are used to process the original data of each pipeline-transported liquid and obtain a standard model for each pipeline-transported liquid:

[0073] F1: Read the built-in reflectivity plate data and the built-in reflectivity plate calibration file during acquisition. Collect several frames of dark background data according to the preset acquisition rules. For example, you can set the data size of each frame to 480*300, where 480 represents the number of pixels and 300 represents the number of channels. Collect 50 frames in total and calculate the average dark background frame data. Preferably, the following formula can be used to calculate the average dark background frame data:

[0074] ;

[0075] in: is the average frame data of dark background, is the dark background data of the mth frame;

[0076] F2: Calculate the corresponding frame reflectivity data and perform transposition to obtain the transposed reflectivity data. In specific implementation, the following formula can be used to calculate the corresponding frame reflectivity data:

[0077] ;

[0078] Where: R is a frame of reflectivity data, A is the original data of the liquid transported in the pipeline, B is the built-in reflectivity plate data, and C is the built-in reflectivity plate calibration file;

[0079] F3: Import the transposed reflectivity data into the preset model for training to obtain a standard model for each type of pipeline transport liquid.

[0080] Example 3:

[0081] See also Figure 3 This embodiment provides a pipe flange liquid leakage detection system based on hyperspectral recognition, and a pipe flange liquid leakage detection method based on hyperspectral recognition according to any of the above embodiments is applied, including:

[0082] Main control system 1 is used for overall control, data reception and data processing, and performs comparative analysis based on the collected spectral data to output relevant information about the leaking liquid;

[0083] Leakage monitoring mechanism 2, used to monitor whether there is liquid leakage in the pipeline flange and output a corresponding early warning signal when liquid leakage occurs in the pipeline flange;

[0084] The hyperspectral acquisition mechanism 3 is used to perform hyperspectral acquisition on the target area to obtain spectral data of the target area;

[0085] The acquisition drive mechanism 4 is used to drive the hyperspectral acquisition mechanism 3 to move and adjust the acquisition angle of the hyperspectral acquisition mechanism 3;

[0086] The leakage monitoring mechanism 2, the acquisition drive mechanism 4 and the hyperspectral acquisition mechanism 3 are all data-connected to the main control system 1. The leakage monitoring mechanism 2 is provided with a monitoring optical path for monitoring whether there is liquid leakage in the pipeline flange. The monitoring optical path is connected to the moving path of the leaking liquid. The leakage monitoring mechanism 2 outputs a corresponding early warning signal when the monitoring optical path is cut off. After receiving the corresponding early warning signal, the hyperspectral acquisition mechanism 3 is driven by the acquisition drive mechanism 4 to move to the target position. When completing the hyperspectral acquisition of the target area suspected of liquid leakage, the hyperspectral acquisition mechanism 3 outputs spectral data to the main control system 1. The main control system 1 determines whether there is liquid leakage in the target area after completing the comparative analysis of the spectral data. When working, a protective cover 5 needs to be set on the outer cover of the leakage monitoring mechanism 2, the hyperspectral acquisition mechanism 3 and the acquisition drive mechanism 4 to prevent them from being excessively interfered with by the external environment.

[0087] See also Figures 4 to 8In a specific implementation, in order to be suitable for pipelines with different installation directions, different leakage monitoring mechanisms 2 need to be adopted. For example, when the pipeline is arranged vertically, the leakage monitoring mechanism 2 includes a first infrared generator 21, a first infrared receiver 22, a polygonal reflector group 23 and a mounting clamp 24. The mounting clamp 24 is arranged around the periphery of the corresponding pipeline flange. The upper inner wall surface of the mounting clamp 24 maintains a gap with the pipeline flange to allow leaking liquid to flow in. The polygonal reflector group 23 is provided with a plurality of polygonal slices 231. The plurality of polygonal slices 231 are fixedly installed on the upper inner wall surface of the mounting clamp 24 in sequence according to preset angles around the pipeline flange. The plurality of polygonal slices 231 are provided with a reflective film for reflecting infrared light. The first infrared generator 21 and the first infrared receiver 2 2 are arranged inside the upper end of the mounting clamp 24, the first infrared generator 21 and the first infrared receiver 22 are arranged opposite to each other and emit infrared light to illuminate the polygonal reflector group 23, and the polygonal reflector group 23 refracts the infrared light to the receiving part of the first infrared receiver 22 to establish a monitoring light path. Preferably, the inner end tops of the several polygonal slices 231 are provided with a guide slope 2311 for guiding the leaking liquid into the gap. When the leaking liquid flows into the mounting clamp 24, the monitoring light path will be interrupted, thereby outputting an early warning signal. As a further improvement of this embodiment, a guide groove 2312 can also be provided at the inner end top of the polygonal slice 231, which can accurately guide the leaking liquid to the irradiation part of the infrared light on the polygonal slice 231, thereby further improving its monitoring accuracy.

[0088] As a matching collection drive mechanism 4, when the pipeline is arranged vertically, the collection drive mechanism 4 includes an annular guide rail 41, a first sliding mounting seat 42 and a first driving device 43. The annular guide rail 41 is arranged around the periphery of the corresponding pipeline flange, and the first sliding mounting seat 42 is slidably limited and mounted on the annular guide rail 41. The output part of the first driving device 43 is transmission-connected with the first sliding mounting seat 42. The hyperspectral collection mechanism 3 is installed on the first sliding mounting seat 42 and moves to the target position when the first driving device 43 drives the first sliding mounting seat 42 to move.

[0089] Of course, in actual work, when the pipeline is arranged horizontally, the leakage monitoring mechanism 2 includes a second infrared generator 25, a second infrared receiver 26 and a mounting rack 27. The second infrared generator 25 and the second infrared receiver 26 are both installed on the mounting rack 27 and arranged relatively below the corresponding pipeline flange. The second infrared generator 25 emits infrared light to illuminate the receiving part of the second infrared receiver 26 to establish a monitoring optical path.

[0090] As a matching collection drive mechanism 4, when the pipeline is arranged horizontally, the collection drive mechanism 4 includes a translation rail 44, a second sliding mounting seat 45 and a second driving device 46. The translation rail 44 is horizontally arranged below the corresponding pipeline flange, and the second sliding mounting seat 45 is slidably limited and mounted on the translation rail 44. The hyperspectral collection mechanism 3 is installed on the second sliding mounting seat 45 and moves to the target position when the second driving device 46 drives the second sliding mounting seat 45 to move.

[0091] During operation, when liquid leaks and drips from the pipe flange, the leaked liquid enters the monitoring optical path, causing the monitoring optical path of the leakage monitoring mechanism 2 to be interrupted. In this embodiment, the leakage monitoring mechanism 2 can send a signal carrying the device number to the main control system 1 via the 4G communication module. The main control system 1 obtains this signal by subscribing to the corresponding MQTT topic, thereby accurately identifying the specific location of the leakage monitoring mechanism 2 that triggered the leak detection. Upon receiving the warning signal, the main control system 1 immediately activates the hyperspectral acquisition mechanism 3 to collect hyperspectral data. The acquisition drive mechanism 4 moves the hyperspectral acquisition mechanism 3 to the target location. After the hyperspectral acquisition mechanism 3 completes hyperspectral data collection in the target area, the main control system 1 pre-processes the collected hyperspectral data. The spectral data is compared and analyzed with a preset standard liquid spectral database to determine whether liquid leakage exists in the target area and whether the detected liquid is the target liquid to be monitored. If the model determines that the detected liquid is the target monitoring liquid, the main control system 1 will push a leak warning message to the email address and mobile phone of the pre-registered responsible person via the MQTT protocol, notifying the relevant personnel to take timely action. At the same time, the main control system 1 will mark the status of the leakage point as leakage, display it in red on the corresponding monitoring screen, and start the sound and light alarm device to remind. When the person in charge starts the maintenance work, he can manually mark the status of the point as under maintenance. In the maintenance state, the main control system 1 will suspend receiving MQTT messages from the point, effectively avoiding false alarm interference caused by maintenance operations. If the model determines that the detected liquid is not the target monitoring liquid, it is regarded as a false alarm. The main control system 1 will discard the message and record the false alarm information. After waiting for the leakage monitoring mechanism 2 to restore the monitoring optical path, it will restart the leakage detection process to complete the full-time monitoring of the pipeline flange.

[0092] The beneficial technical effects of this embodiment include: the present invention can realize a series of tasks such as full-time monitoring of pipeline flanges, liquid leakage monitoring and early warning, liquid leakage monitoring and determination work and liquid sufficiency intelligent judgment, with a high degree of automation and intelligence, reducing the proportion of manual participation, and high real-time and accurate detection. It can effectively improve the overall detection efficiency and accuracy, and realize real-time and accurate detection of tiny leakage points and intelligent judgment of liquid composition in complex industrial environments.

[0093] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Those skilled in the art will understand that the present invention includes, but is not limited to, the contents described in the drawings and the above specific embodiments. Any modifications that do not deviate from the functional and structural principles of the present invention are intended to be included within the scope of the claims.

Claims

1. A method for detecting liquid leakage in pipe flanges based on hyperspectral recognition, characterized in that: The following steps are involved: S1: Initialize the pipeline flange liquid leakage detection system and emit a monitoring light beam to establish a monitoring light path for monitoring whether there is liquid leakage in the pipeline flange. When liquid leakage occurs in the pipeline flange and the monitoring light path is cut off, the system proceeds to step S2; S2: Perform preliminary fluid leakage localization to locate the target area where fluid leakage is suspected; S3: Performing hyperspectral acquisition on the target area to obtain spectral data of the target area, comparing and analyzing the spectral data with a preset standard liquid spectrum database to determine whether there is liquid leakage in the target area. If there is liquid leakage, relevant information of the leaked liquid is output synchronously; In step S3, the spectral data is compared with a preset standard liquid spectral database to determine whether there is liquid leakage in the target area. When liquid leakage is present, the relevant information of the leaked liquid is output synchronously, and the following steps are adopted: E1: Calculate the reflectance data of the spectral data, perform first-order derivative on each pixel point of a frame of spectral data, and obtain a first-order derivative data frame; E2: Compare the first-order derivative curve of each pixel in the first-order derivative data frame with the curve recorded in the standard liquid spectrum database to determine whether the spectral data matches the curve recorded in the standard liquid spectrum database. A variance value is obtained through Euclidean calculation. When the variance value is less than the set threshold, it is considered that this curve is similar to the corresponding standard sample. When the variance value is similar to multiple curves, the curve with the smallest variance value is selected as the final result and the matching result is output.

2. The method for detecting liquid leakage of pipeline flanges based on hyperspectral recognition according to claim 1, characterized in that: In step S1, the pipeline flange liquid leakage detection system is initialized and a monitoring light beam is emitted to establish a monitoring light path for monitoring whether there is liquid leakage in the pipeline flange. The following steps are adopted: A1: The leakage monitoring mechanism (2) provided in the pipeline flange liquid leakage detection system is arranged around the periphery of the pipeline flange, so that the reflection portion provided in the leakage monitoring mechanism (2) is arranged around the periphery of the pipeline flange; A2: The leakage monitoring mechanism (2) is provided with an infrared generating device to emit infrared light along a preset angle. The infrared light is irradiated on the reflecting part and refracted in sequence to form an infrared light path arranged around the pipe flange. The leakage monitoring mechanism (2) is provided with an infrared receiving device to be arranged at the end of the infrared light path to receive the infrared light, thereby establishing a monitoring light path for monitoring whether there is liquid leakage in the pipe flange.

3. The method for detecting liquid leakage of pipeline flanges based on hyperspectral recognition according to claim 1, characterized in that: In step S1, the pipeline flange liquid leakage detection system is initialized and a monitoring light beam is emitted to establish a monitoring light path for monitoring whether there is liquid leakage in the pipeline flange. The following steps are adopted: B1: Arrange the leakage monitoring mechanism (2) provided in the pipeline flange liquid leakage detection system around the periphery of the pipeline flange; B2: The infrared generating device and the infrared receiving device are arranged relative to each other in the leakage monitoring mechanism (2), and the infrared generating device reflects the infrared light to establish an infrared light path, and the infrared light path is connected to the movement path of the leaking liquid to establish a monitoring light path for monitoring whether there is liquid leakage in the pipeline flange.

4. The method for detecting liquid leakage of pipeline flanges based on hyperspectral recognition according to claim 1, characterized in that: In the step S3, the following steps are also included: after calling the pre-collected pipeline model data to perform background stripping on the spectral data of the target area, the spectral data is compared and analyzed with a preset standard liquid spectral database to determine whether there is liquid leakage in the target area.

5. The method for detecting liquid leakage of pipeline flanges based on hyperspectral recognition according to claim 1, characterized in that: The establishment of the standard liquid spectrum database adopts the following steps: D1: Obtain several types of pipeline transported liquids and collect the original data of each pipeline transported liquid; D2: Processing the raw data of each pipeline transported liquid to obtain a standard model for each pipeline transported liquid; D3: Obtain relevant information for several practical application scenarios, assign weights to several standard models, and store them as a standard liquid spectrum database.

6. A pipe flange liquid leakage detection system based on hyperspectral recognition, applying the pipe flange liquid leakage detection method based on hyperspectral recognition as described in any one of claims 1 to 5, characterized in that: include: The main control system (1) is used for overall control, data reception and data processing, and performs comparative analysis based on the collected spectral data to output relevant information about the leaked liquid; A leakage monitoring mechanism (2) is used to monitor whether there is liquid leakage in the pipeline flange and output a corresponding early warning signal when liquid leakage occurs in the pipeline flange; A hyperspectral acquisition mechanism (3) is used to perform hyperspectral acquisition on a target area to obtain spectral data of the target area; An acquisition drive mechanism (4) is used to drive the hyperspectral acquisition mechanism (3) to move and adjust the acquisition angle of the hyperspectral acquisition mechanism (3); The leakage monitoring mechanism (2), the acquisition drive mechanism (4) and the hyperspectral acquisition mechanism (3) are all data-connected to the main control system (1). The leakage monitoring mechanism (2) is provided with a monitoring light path for monitoring whether there is liquid leakage in the pipeline flange, and the monitoring light path is connected to the movement path of the leaking liquid. The leakage monitoring mechanism (2) outputs a corresponding early warning signal when the monitoring light path is cut off. After receiving the corresponding early warning signal, the hyperspectral acquisition mechanism (3) is driven by the acquisition drive mechanism (4) to move to the target position. When the hyperspectral acquisition mechanism (3) completes the hyperspectral acquisition of the target area suspected of having liquid leakage, it outputs spectral data to the main control system (1). The main control system (1) determines whether there is liquid leakage in the target area after completing the comparative analysis of the spectral data.

7. The pipeline flange liquid leakage detection system based on hyperspectral recognition according to claim 6 is characterized by: The leakage monitoring mechanism (2) comprises a first infrared generator (21), a first infrared receiver (22), a polygonal reflector group (23) and a mounting hoop (24), wherein the mounting hoop (24) is arranged around the periphery of the corresponding pipe flange, and an inner wall surface of the upper end of the mounting hoop (24) maintains a gap with the pipe flange to allow leakage liquid to flow in, and the polygonal reflector group (23) is provided with a plurality of polygonal slices (231), and the plurality of polygonal slices (231) are fixedly mounted in sequence around the pipe flange at a preset angle. On the inner wall surface of the upper end of the mounting clamp (24), a plurality of polygonal slices (231) are provided with reflective films for reflecting infrared light. The first infrared generator (21) and the first infrared receiver (22) are both arranged inside the upper end of the mounting clamp (24). The first infrared generator (21) and the first infrared receiver (22) are arranged opposite to each other and emit infrared light to illuminate the polygonal reflector group (23). The polygonal reflector group (23) refracts the infrared light to the receiving part of the first infrared receiver (22) to establish a monitoring light path.

8. The pipeline flange liquid leakage detection system based on hyperspectral recognition according to claim 7 is characterized in that: The acquisition drive mechanism (4) comprises an annular guide rail (41), a first sliding mounting seat (42) and a first driving device (43). The annular guide rail (41) is arranged around the periphery of the corresponding pipeline flange. The first sliding mounting seat (42) is slidably mounted on the annular guide rail (41). The output portion of the first driving device (43) is transmission-connected to the first sliding mounting seat (42). The hyperspectral acquisition mechanism (3) is mounted on the first sliding mounting seat (42) and is displaced to a target position when the first driving device (43) drives the first sliding mounting seat (42) to operate.

9. The pipeline flange liquid leakage detection system based on hyperspectral recognition according to claim 6 is characterized in that: The leakage monitoring mechanism (2) comprises a second infrared generator (25), a second infrared receiver (26) and a mounting frame (27). The second infrared generator (25) and the second infrared receiver (26) are both mounted on the mounting frame (27) and arranged relative to each other below the corresponding pipe flange. The second infrared generator (25) emits infrared light to illuminate the receiving portion of the second infrared receiver (26) to establish a monitoring light path.

10. The pipeline flange liquid leakage detection system based on hyperspectral recognition according to claim 9 is characterized in that: The acquisition drive mechanism (4) comprises a translation rail (44), a second sliding mounting seat (45) and a second drive device (46). The translation rail (44) is horizontally arranged below the corresponding pipeline flange. The second sliding mounting seat (45) is slidably mounted on the translation rail (44). The hyperspectral acquisition mechanism (3) is mounted on the second sliding mounting seat (45) and moves to a target position when the second drive device (46) drives the second sliding mounting seat (45) to move.

Images