An eccentric butterfly valve performance detection method and system based on optical detection

Abstract

The application provides a kind of eccentric butterfly valve performance detection method and system based on optical detection, the method comprises: obtaining the first image of the light transmission intensity distribution of the sealing pair structure of bidirectional full-metal hard sealing four eccentric butterfly valve;According to the reflection interference image of structured light on the sealing pair structure collected by phase modulation structure grating light field, and combined with the first image, joint analysis is carried out to generate contact pressure distribution matrix;Based on the first image and contact pressure distribution matrix, the corresponding multi-dimensional image features are extracted to generate leakage early warning area atlas;The floating response trajectory of the sealing pair structure in the opening and closing process is compared with the leakage early warning area atlas in space, and the corresponding floating response characteristics of the leakage early warning area in the opening and closing process are determined;According to the abnormal degree of floating response characteristics, the butterfly valve sealing performance detection result is generated. Through the implementation of the application scheme, the leakage early warning area and its corresponding dynamic abnormal characteristics can be identified in advance, and the safety and operation stability of the system are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of valves, in particular to a performance detection method and system of an eccentric butterfly valve based on optical detection. BACKGROUND

[0002] In typical industrial scenarios such as petroleum and chemical industry, coal chemical industry and high temperature and high pressure steam pipe network, bidirectional full metal hard sealing four eccentric butterfly valves are widely used for high reliability control of pressure-bearing medium. Due to its structural characteristics suitable for bidirectional pressure difference, large flow scouring and high temperature corrosion and other harsh working conditions, its sealing performance is directly related to the safety and stability of the system. However, the current sealing performance detection means of such butterfly valves mainly rely on leakage rate measurement or morphology observation after disassembly. Not only is the detection lagging and costly, but it is also difficult to accurately reflect the real contact state and dynamic response behavior of the sealing pair structure during opening and closing, especially it is unable to identify potential local failure areas or structural deviations in advance. Especially in the full metal sealing pair structure, the contact pressure change caused by slight uneven fitting, assembly deviation or wear is often not enough to cause short-term leakage, but it may cause sudden leakage in high-frequency opening and closing or pressure difference fluctuation. SUMMARY

[0003] The present application provides a performance detection method and system of an eccentric butterfly valve based on optical detection, which solves the problem that related technologies cannot identify potential local failure areas or structural deviations in advance.

[0004] The first aspect of the present application provides a performance detection method of an eccentric butterfly valve based on optical detection, which comprises:

[0005] obtaining a first image of the light transmission intensity distribution of the sealing pair structure of a bidirectional full metal hard sealing four eccentric butterfly valve;

[0006] collecting the reflection interference image of the structured light on the sealing pair structure according to the phase modulation structured grating light field acquisition, and combining the first image for joint analysis to generate a contact pressure distribution matrix;

[0007] based on the first image and the contact pressure distribution matrix, extracting corresponding multi-dimensional image features to generate a leakage warning area atlas;

[0008] spatially comparing the leakage warning area atlas with the floating response trajectory of the sealing pair structure in the opening and closing process to determine the corresponding floating response features of the leakage warning area in the opening and closing process;

[0009] generating a butterfly valve sealing performance detection result according to the abnormality degree of the floating response features.

[0010] The second aspect of the application provides an optical detection-based eccentric butterfly valve performance detection system for realizing an optical detection-based eccentric butterfly valve performance detection method.

[0011] An acquisition module is configured to acquire a first image of light transmission intensity distribution of a contact area between a sealing pair structure of a bidirectional full-metal hard sealing four-eccentric butterfly valve;

[0012] An analysis module is configured to acquire a reflection interference image of structured light on the sealing pair structure according to a phase modulation structure grating light field acquisition structure light, and perform joint analysis in combination with the first image to generate a contact pressure distribution matrix;

[0013] A construction module is configured to extract corresponding multi-dimensional image features to generate a leakage early warning area atlas based on the first image and the contact pressure distribution matrix;

[0014] A determination module is configured to perform spatial comparison between the leakage early warning area atlas and a floating response track of the sealing pair structure in an opening and closing process, and determine corresponding floating response features of the leakage early warning area in the opening and closing process.

[0015] A generation module is configured to generate a butterfly valve sealing performance detection result according to an abnormality degree of the floating response features.

[0016] The third aspect of the embodiments of the application provides an electronic device, including a memory and a processor, wherein the processor is configured to execute a computer program stored in the memory, and when the processor executes the computer program, each step of the optical detection-based eccentric butterfly valve performance detection method provided in the first aspect of the embodiments of the application is implemented.

[0017] The fourth aspect of the embodiments of the application provides a computer readable storage medium, which stores a computer program, and when the computer program is executed by a processor, each step of the optical detection-based eccentric butterfly valve performance detection method provided in the first aspect of the embodiments of the application is implemented.

[0018] In summary, according to the method and system for detecting performance of an eccentric butterfly valve based on optical detection provided in the application, a first image of light transmission intensity distribution of a sealing pair structure of a bidirectional full-metal hard sealing four-eccentric butterfly valve is acquired; a reflection interference image of a structured light on the sealing pair structure is collected according to a phase modulation structured light grating, and joint analysis is performed in combination with the first image to generate a contact pressure distribution matrix; corresponding multi-dimensional image features are extracted based on the first image and the contact pressure distribution matrix to generate a leakage warning area atlas; the leakage warning area atlas is compared with a floating response trajectory of the sealing pair structure in an opening and closing process in space to determine corresponding floating response features of the leakage warning area in the opening and closing process; and a butterfly valve sealing performance detection result is generated according to an abnormality degree of the floating response features. Through implementation of the application, the leakage warning area and corresponding dynamic abnormal features can be identified in advance, and the safety and operation stability of the system are improved. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 A flowchart of the method for detecting performance of an eccentric butterfly valve based on optical detection provided in the embodiment of the application is shown in the figure.

[0020] Figure 2 A program module schematic diagram of the system for detecting performance of an eccentric butterfly valve based on optical detection provided in the embodiment of the application is shown in the figure.

[0021] Figure 3 A structure schematic diagram of the electronic device provided in the embodiment of the application is shown in the figure. DETAILED DESCRIPTION

[0022] In order to make the application purposes, features and advantages of the application more obvious and easy to understand, the technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the application.

[0023] In order to solve the problem that the related art cannot identify potential local failure areas or structural deviations in advance, the embodiment of the application provides a method for detecting performance of an eccentric butterfly valve based on optical detection, as shown in the figure. Figure 1 A flowchart of the method for detecting performance of an eccentric butterfly valve based on optical detection provided in the embodiment is shown in the figure. The method for detecting performance of an eccentric butterfly valve based on optical detection includes the following steps:

[0024] In step 110, a first image of light transmission intensity distribution of a contact area between sealing pair structures of a bidirectional full-metal hard sealing four-eccentric butterfly valve is acquired.

[0025] In the present embodiment, in order to evaluate the contact state of the sealing pair structure (referring to the contact surface composed of the sealing ring of the disc valve and the metal valve seat, which is the metal-to-metal sealing pair structure of the four-eccentric butterfly valve) of the bidirectional full-metal hard sealing four-eccentric butterfly valve, a first image of the light transmission intensity distribution of the contact area between the sealing ring of the disc valve and the metal valve seat needs to be obtained. The image is obtained by backlight transmission imaging of the sealing contact surface when the disc valve and the valve seat are in the closed state. The image can reflect the optical transmission rate difference of the contact area, indirectly reveal the fitting degree and micro-gap distribution between the sealing pairs, and thus provide a reference for the optical characteristics of the static sealing surface for subsequent analysis. In addition, the first image can also be used as a reference for constructing the subsequent coordinate reference and analysis mask to improve the spatial positioning accuracy of the interference analysis and speckle tracking.

[0026] Step 120: collecting a reflection interference image of the structured light on the sealing pair structure according to the phase-modulated structured light grating light field, and performing joint analysis combined with the first image to generate a contact pressure distribution matrix.

[0027] In the present embodiment, by projecting a structured light grating light field with phase modulation characteristics onto the sealing pair structure, a reflection interference image of the light field on the surface of the sealing pair structure is collected. During the projection of the structured light, the phase parameters of the light field are adjusted so that the interference fringes have a demodulable feature. Subsequently, the reflection interference image and the aforementioned transmission image are jointly analyzed to extract and quantify the local density variation, phase distortion and fringe orientation of the interference fringes, and the interference profile topography of the sealing interface is reconstructed by phase recovery technology. Combined with the transmission difference information of the contact area in the light transmission image, the stress distribution change of the contact interface can be mapped, thereby generating a contact pressure distribution matrix as a quantitative indicator of the micro-contact state of the sealing pair.

[0028] In some embodiments, the step of collecting a reflection interference image of the structured light on the sealing pair structure according to the phase-modulated structured light grating light field, and performing joint analysis combined with the first image to generate a contact pressure distribution matrix, includes: obtaining a light transmission intensity spatial distribution map of the sealing pair structure in the closed state from the first image, and establishing a corresponding initial spatial reference coordinate system of the sealing pair structure; generating a grating light field with a preset phase modulation parameter by controlling the structured light projection unit, and projecting the grating light field onto the sealing pair structure according to the spatial reference coordinate system to collect a reflection interference image sequence of the structured light on the sealing pair structure; extracting the fringe density distribution and reconstructing the phase from the reflection interference image sequence, and combining the light transmission intensity variation characteristics of the same spatial area in the first image to generate a composite interference profile map of the sealing contact interface; and generating a contact pressure distribution matrix according to the phase variation gradient of the interference fringes and the spatial difference mapping relationship of the transmission intensity in the composite interference profile map.

[0029] Specifically, when the butterfly valve is in a closed state, a first image reflecting the micro-adhesion state of the contact area of the sealing pair can be obtained by collecting the transmission light image between the sealing pair structure through a high-resolution optical imaging system. The image takes the spatial transmission intensity as the gray reference, where the sealing tight area appears as low light transmission (low gray value), and the micro-gap or poor adhesion area appears as transmission enhancement (high gray value). By performing two-dimensional spatial gray distribution statistics on the image and setting image boundary control points, a spatial reference coordinate system with the geometric center of the butterfly plate as the origin can be established. This coordinate system is not only used to describe the static spatial structure of the sealing pair structure, but also provides accurate coordinate registration basis for subsequent structured light projection and interference analysis. On the basis of the spatial reference coordinate system, the control structured light projection unit outputs a grating light field with a set phase modulation parameter, where the grating light field is composed of a group of coherent light rays with periodic intensity variation, and the spatial frequency and phase shift amount determine the analytical precision of the subsequent interference fringes. By adjusting the incident angle, phase encoding period and offset modulation amount of the light field, multiple reflection interferences are generated after the light field interacts with the metal reflecting surface of the butterfly plate sealing ring. The projection direction needs to be matched with the sealing curved surface of the butterfly plate through an angle scanning device to realize automatic collimation, thereby ensuring that the light field acts on the effective sealing area. In actual application, a structured grating with a laser wavelength of 532 nm and a grating period of 30 μm is used, and a high-contrast reflection image sequence is generated by adjusting the phase delay to π / 2 for four-step phase shift interference collection. The reflection interference image sequence formed by the structured light on the sealing pair is processed, the dense area and phase abrupt point of the interference fringes in each frame of image are identified by using the fringe density distribution extraction algorithm, and the fringe phase map on the entire sealing surface is reconstructed by using the phase unwrapping and filtering recovery technology. Since the interference fringe density of different areas reflects the spatial variation of the optical path difference, the micro-height distribution on the sealing surface can be restored. Combined with the light transmission intensity characteristics of the corresponding area in the first image, i.e. the relationship between the gray value and the optical path difference, the spatial joint analysis of the double-source data can be realized. The joint analysis can correct the interference map error caused by uneven illumination or local reflectivity difference, and improve the stability and spatial alignment accuracy of the interference fringe demodulation. The composite interference contour generated thereby simultaneously represents the topography fluctuation and the distribution characteristics of the local transmission flux between the metal sealing pair in space, and can reveal the micro-adhesion state of the sealing contact surface under the premise of good macro-geometric cooperation. Especially in the case that the sealing ring is locally deformed, the metal build-up is uneven, or the residual stress of the weld causes adhesion distortion, the image can display obvious fringe distortion and transmission enhancement area, which prompts the potential micro-permeation risk. Finally, the mapping relationship between the phase variation gradient of the fringes in the composite interference contour map, i.e. the change rate of the fringe phase value per unit distance, and the spatial distribution of the light transmission intensity can be established, so as to calculate the local contact pressure of each sealing contact unit.Physically, the area with more severe interference phase change usually corresponds to larger local deformation and weaker contact force on the contact surface, while the enhanced transmission intensity further indicates the insufficient contact in this area. By fitting the phase gradient value and the transmission gray value, a mathematical mapping model between the value and the contact pressure per unit area is established, and finally a contact pressure distribution matrix is generated. The distribution matrix is a two-dimensional array structure, where each element represents the unit pressure value of the butterfly plate sealing ring on the metal valve seat at a certain spatial point in the closed state, thereby providing a physical basis for judging whether the contact is sufficient and the sealing is reliable. The distribution matrix can be used in defect modeling, dynamic response analysis, and structure compensation design, etc.

[0030] Step 130, based on the first image and the contact pressure distribution matrix, extracting corresponding multi-dimensional image features to generate a leakage warning area map.

[0031] In this embodiment, based on the obtained first image and contact pressure distribution matrix, image features of multiple dimensions are extracted by image feature extraction algorithm, such as gray gradient, local structure tensor, light transmission boundary sharpness, interference texture frequency response, etc. Using these features, a label relationship between image features and defect distribution is established, and a defect label classification model is constructed. Through the fusion processing of multi-image channel information, this model can identify the potential micro-defect position in the sealing pair structure, including micro-leakage seam, uneven contact area and contact stress mutation area, etc., and form a spatial distribution map based on the defect position and risk intensity, and finally generate a leakage warning area map as a graphical identification of potential leakage source in the sealing pair structure.

[0032] In some embodiments, based on the first image and the contact pressure distribution matrix, the step of extracting corresponding multi-dimensional image features to generate a leakage warning area map includes: performing spatial registration on the pixel coordinate relationship of the corresponding areas of the first image and the contact pressure distribution matrix to generate a composite image feature set with unified spatial index; extracting a multi-dimensional structure feature vector representing abnormal sealing behavior by jointly statistical the composite sealing features of each sub-area in the composite image feature set; inputting the multi-dimensional structure feature vector into a preset defect label classification model to classify the defects in each sub-area of the composite image feature set and determine the corresponding defect label classification result; generating a defect distribution map of the sealing pair structure according to the defect label classification result, and marking the areas in the defect distribution map of the sealing pair structure with defect level higher than a preset threshold as leakage warning areas to generate a leakage warning area map.

[0033] Specifically, the first image records the light transmission intensity distribution of the seal pair structure in the closed state, and the contact pressure distribution matrix obtains the microscopic pressure information of the same region through the interference fringe phase reconstruction. Due to the differences in imaging parameters and spatial resolution due to different sources, a unified reference frame must be established through pixel coordinate mapping. The registration process can map each physical point in the contact pressure matrix to the gray pixel position of the first image by means of image transformation algorithms such as affine transformation or homography mapping based on feature points, and finally generate a composite image feature set represented by a unified spatial index. Each unit of this feature set is located by a set of spatial coordinates and carries composite attribute information from different image sources, such as gray intensity, local gradient, contact pressure, transmission rate change, etc. Based on this composite image feature set, by dividing multiple spatial sub-regions and performing local statistical processing in each region, composite seal features are extracted. This feature not only includes the original gray and pressure values, but also covers its spatial first-order gradient, second-order partial derivative, variance, mean shift, structural entropy, and local edge response. These parameters collectively depict the fitting state, stress concentration trend, and topographic mutation characteristics of the seal pair structure surface at macro and micro scales. The significance of multi-dimensional statistics is to enhance the recognizability of local abnormal behavior, especially in scenarios where the interface structure is continuous but has potential micro-defects. For example, a region with uniform gray but pressure mutation is not easily detected in an independent image, and the statistical joint feature significantly improves the signal-to-noise ratio of its abnormal performance. Subsequently, the multi-dimensional structural feature vectors of all sub-regions are reconstructed, the feature space is compressed and reorganized through principal component analysis, linear discriminant analysis or auto-encoding network, and the abnormal sensitive factor is calibrated combined with the label information of known defect regions in historical data. This calibration process trains a defect label classification model through a supervised learning mechanism, and the model type can use support vector machines, convolutional neural networks or graph convolution networks to identify the corresponding structure state category according to the input feature vector, including "no defect", "mild poor fit", "local separation" and "high-risk penetration zone" labels. In the identification process, the model performs non-linear mapping on the feature vector and outputs the classification probability, forming the defect label classification result for each sub-region. According to the above classification result, a seal pair structure defect distribution map is further generated, combining the classification labels and spatial positions of all sub-regions to form a visual two-dimensional defect distribution image. This image shows the defect distribution pattern, severity level and spatial aggregation trend of the entire seal surface. To further strengthen risk identification, the areas with defect levels higher than the preset threshold in the distribution map are highlighted, their boundary contours are extracted and assigned special warning labels, thereby constructing a leakage warning area map.The atlas reflects high-risk penetration areas based on multi-source image fusion and structural learning analysis, which can not only be used for seal pair performance evaluation, but also serve as input basis for subsequent floating response trajectory matching analysis, and provide spatial reference for dynamic behavior anomaly identification. For example, there is a machining burr residual area at the edge of the outer circle of the butterfly valve, which is accurately identified as a warning area in the atlas and highly consistent with the actual leakage position, verifying the effectiveness of the technical path in seal risk prediction.

[0034] In some embodiments, the opening and closing of the seal pair structure is controlled according to a preset opening and closing time period, and an image sequence of a preset optical speckle pattern of the seal pair structure in consecutive time frames is collected during the opening and closing process; by sub-pixel registration of the optical speckle pattern between adjacent time frames in the image sequence, the displacement of the optical speckle pattern in the spatial coordinate system is extracted, and a sequence of speckle displacement vectors is generated; by time dimension slicing analysis of the sequence of speckle displacement vectors, a surface displacement field of the seal ring corresponding to each time frame is constructed, and according to the spatial displacement continuity and amplitude distribution characteristics, a local dynamic response feature map of each time frame is determined; according to the spatiotemporal continuous distribution of the displacement change abnormal area in the local dynamic response feature map, the high response trajectories of each time frame are tracked and superimposed, and the floating response trajectory of the seal ring of the butterfly plate in the complete opening and closing period is generated.

[0035] Specifically, in the present embodiment, the relative motion between the sealing pair structure is precisely controlled according to the preset opening and closing time period, and at the same time, the image sequence of the sealing ring surface changing with time is recorded by the high frame rate image acquisition system. The image acquisition area needs to be sprayed or etched with a random distribution of optical speckle pattern in advance. The speckle pattern can be realized by spray coating, laser etching or micro-nano pattern transfer, etc. Its role is to provide tracking reference points for subsequent displacement analysis, especially in the absence of metal surface texture or high light environment, the strong interference of artificial speckle structure can significantly improve the image registration accuracy. In order to extract the actual displacement of the speckle pattern in the image sequence, sub-pixel level image registration is required between adjacent time frames. Sub-pixel registration refers to higher precision displacement estimation of the speckle pattern position in the image through sub-pixel interpolation algorithm, for example, phase correlation technology based on cross-correlation. In this process, the system divides the speckle pattern area in each frame into several overlapping or sliding search windows, and accurately calculates the matching displacement between the front and rear frames for each window, finally forming a two-dimensional vector field, each vector represents the relative displacement of the speckle pattern at that position. This displacement vector sequence completely records the spatial behavior of the micro-deformation and local floating of the sealing pair structure in the entire opening and closing period. Next, by slicing the speckle displacement vector sequence in the time dimension, the sealing ring surface displacement field corresponding to each time frame can be constructed. The displacement field takes displacement amplitude and direction as basic attributes, which is used to describe the motion intensity and direction trend of the local area at different time points. In order to improve the physical interpretability of displacement field analysis, it is necessary to combine spatial continuity indicators (such as displacement gradient field, local smoothness) and amplitude variation characteristics (such as instantaneous jump rate, adjacent area difference statistics) for response behavior analysis, so as to identify the areas with significant dynamic variation characteristics. The obtained local dynamic response feature map marks the abnormal displacement behavior area in each time frame, such as local bulging, displacement anomaly band caused by micro-cracks, stretching spot area caused by stress concentration, etc. According to the spatiotemporal continuous distribution law of the abnormal area in the above local dynamic response feature map, the evolution trajectory of the abnormal area in the entire opening and closing period is tracked, and the floating response trajectory of the butterfly plate sealing ring in the complete opening and closing process is constructed by spatial superposition. The floating response trajectory represents the motion path and structure activity distribution of the high response area on the time axis, which not only contains the motion intensity, but also reflects the hysteresis, frequency disturbance and spatial expansion trend of the response area. For example, in a certain opening and closing process, it is found that a certain area on the sealing ring continuously rises at the beginning of valve opening but quickly falls at the closing time, which may correspond to the initial deformation relaxation and the resulting asymmetric behavior of the structure. The establishment of the floating response trajectory provides dynamic dimension input for the comparison with the leakage warning area map, making the complete behavior identification of the sealing pair structure have the basis of space-time unity, and providing key basis for early leakage trend judgment and structure optimization design.

[0036] Step 140, spatially compare the leakage warning area atlas with the floating response trajectory of the sealing pair structure in the opening and closing process, and determine the corresponding floating response characteristics of the leakage warning area in the opening and closing process.

[0037] In the present embodiment, the leakage warning area atlas is spatially registered and compared with the floating response trajectory, and the corresponding floating response characteristics are extracted according to the displacement response mode and vector direction characteristics of each leakage warning area in the opening and closing process. By statistically analyzing the abnormal displacement, non-continuous deformation or high-frequency disturbance of different warning areas in the dynamic response, the structural response stability of these areas in the valve action process can be determined, so as to reveal whether they have the risk of structural deformation driven leakage. The comparison and analysis make the static defect judgment and dynamic response characteristics associated, and provide multi-angle support for the abnormal source.

[0038] In some embodiments, according to the spatial coordinate information of each defect area in the leakage warning area atlas and the space-time position sequence of each response path in the floating response trajectory, a two-dimensional space mapping relationship of the sealing pair structure is established, and a comparison reference coordinate system is constructed; by spatially dividing the displacement field of each time frame in the floating response trajectory, the displacement sub-area corresponding to the warning area in the reference coordinate system is extracted, and the local response trajectory group matched with each leakage warning area is generated; by jointly analyzing the response directionality, displacement amplitude fluctuation rate and response disturbance mode of the continuous time frames in the local response trajectory group, the dynamic response characteristic vector corresponding to the leakage warning area is extracted; according to the time sequence evolution law and spatial aggregation degree of each characteristic parameter in the dynamic response characteristic vector, the floating response characteristics of each leakage warning area in the opening and closing process are determined.

[0039] Specifically, in the present embodiment, in order to realize the identification of the coupling between the leakage risk area in the sealing pair structure and its dynamic response behavior, first, the spatial mapping relationship between the spatial coordinate information of each defect area in the leakage early warning area atlas and the spatial position sequence of the response path contained in the floating response trajectory on the time axis is established, thereby constructing a unified comparison reference coordinate system. The reference coordinate system serves as a basic framework for spatial alignment, used to coordinate the regional consistency of static atlas and dynamic trajectory in the spatial domain. In the construction process, affine transformation or non-rigid transformation model based on feature point registration is introduced, so that the static defect areas (such as cracks, peeling or stress unevenness) in the sealing surface image and the space-time response units in the floating response trajectory can be correspondingly positioned in the same two-dimensional coordinate system. For example, if an arc-shaped high-level defect area is marked in the leakage early warning area atlas, this area should be projected onto the dynamic displacement trajectory with a similar spatial path in the floating response trajectory. Subsequently, under the unified reference coordinate system, the displacement field corresponding to each time frame in the floating response trajectory is subjected to spatial segmentation processing, thereby extracting the local displacement sub-area consistent with the position of the leakage early warning area, and based on the time sequence integration, forming a local response trajectory group corresponding to each early warning area. The response trajectory group takes the time evolution path of all response points in the region as a set, retaining the deformation trend and response fluctuation behavior of the local structure in the opening and closing period. The spatial segmentation process can use a mask matching algorithm, and the corresponding mask is generated from the defect area outline in the leakage early warning area atlas, ensuring that the intercepted response trajectory has spatial consistency. For example, for an elliptical crack area, the system will divide this area in the time-continuous displacement field into an independent trajectory set for analyzing its structural response characteristics. By deeply analyzing the response behavior between the time-continuous frames in each local response trajectory group, a dynamic response feature vector reflecting the mechanical instability of the defect area can be extracted. The feature vector integrates multiple response dimensions, including the angle offset of response directionality, the fluctuation rate of displacement amplitude, and the frequency domain structure of response disturbance pattern. Response directionality refers to the change of displacement vector direction in consecutive frames, which can reflect the consistency or abruptness of structural deformation direction; the amplitude fluctuation rate measures the degree of change in displacement intensity, reflecting whether there is periodic contraction or sudden relaxation; the disturbance pattern can extract the frequency components of local response through short-time Fourier transform, identifying high-frequency disturbance or low-frequency drift behavior. For example, in a certain defect area, if the dynamic response feature vector shows that the displacement direction frequently reverses, the amplitude is unstable, and there is a disturbance frequency in the 5-10 Hz interval, it can be initially judged that the area may have contact gap instability or local structure loosening phenomenon. Finally, through the joint analysis of the time sequence evolution rule and the spatial aggregation degree of the extracted dynamic response feature vector, the specific floating response characteristics of each leakage early warning area in the opening and closing period can be determined.The time evolution law mainly investigates the trend form of the characteristic parameters with time, such as whether the response amplitude presents an increasing, decreasing or oscillatory evolution, and the spatial aggregation degree measures whether the high response value is concentrated in the inside or the edge of the early warning area. If there is a persistent high response point aggregation phenomenon in the inside of a certain early warning area, and the characteristic vector presents a periodic oscillation and amplitude amplification trend on the time axis, it can be inferred that the floating response characteristics of the region show the repeated opening and closing fluctuation behavior induced by stress asymmetry. This floating response characteristic not only reveals the mechanical activity of the potential leakage area, but also provides data support and characteristic basis for subsequent sealing performance level evaluation and structure optimization.

[0040] Step 150, generating the butterfly valve sealing performance detection result according to the abnormal degree of the floating response characteristics.

[0041] In this embodiment, the abnormal degree of the floating response characteristics is combined with the defect probability output by the defect label classification model, and the response amplitude change rate, spatial discontinuity and defect saliency are jointly weighted to form a comprehensive evaluation basis for sealing performance. Through the joint model to generate the butterfly valve sealing performance detection result, the overall sealing state level, potential leakage trend prediction and structure response integrity evaluation can be output, realizing the quantitative, early and structure-related detection goal of sealing performance.

[0042] In some embodiments, a multi-dimensional response abnormality index distribution map is generated according to the dynamic response abnormality characteristic parameters of each response area in the floating response characteristics; by spatial superposition analysis of the multi-dimensional response abnormality index distribution map and the defect probability distribution map output by the defect label classification model, the joint distribution characteristics of the response abnormality index and the defect probability in each spatial sub-area are extracted, and a sealing pair structure cooperative abnormality recognition map is generated; according to the joint weight relationship of the response abnormality index and the defect probability in the sealing pair structure cooperative abnormality recognition map, a sealing area grading evaluation rule is established, and the sealing performance level of each sub-area is divided to determine the corresponding local performance level label; through spatial clustering and level weighting statistics of all local performance level labels, the butterfly valve overall sealing performance evaluation parameters are obtained, and the butterfly valve sealing performance detection result is output.

[0043] Specifically, to achieve comprehensive performance evaluation of the seal pair structure, a multi-dimensional response anomaly index distribution map is constructed based on the dynamic response anomaly characteristic parameters of each response region in the floating response feature. The dynamic response anomaly characteristic parameters include response direction disturbance amplitude, displacement vector change rate, response frequency disturbance component, and response path discontinuity, etc. These parameters are all derived from the floating response trajectory captured by the seal ring during the opening and closing cycle. By quantifying and normalizing the above parameters in the spatial dimension and mapping them to a unified coordinate system, an index map reflecting the degree of response anomaly can be generated on the sealing surface. The map takes each response region as a pixel unit and outputs the dynamic instability intensity exhibited in the time evolution process. For example, in a high-frequency oscillation region of the seal ring, the response direction changes dramatically and the displacement speed increases suddenly, so the corresponding anomaly index value will be significantly higher than that of the surrounding area, forming a high-light abnormal identification in the index distribution map. In order to further identify the synergistic abnormal regions of the seal pair structure that have both response anomalies and structural defects, a spatial superposition analysis of the multi-dimensional response anomaly index distribution map and the defect probability distribution map output by the defect marker classification model is required. The defect probability distribution map is output by the image recognition model constructed in the early stage, representing the probability level of the existence of surface structural defects (such as cracks, corrosion, and local depressions) in each spatial sub-region. By performing pixel-level registration processing on the two images in space, the response anomaly index value and the defect probability value corresponding to each sub-region can be obtained in the same spatial coordinate system, and then the joint distribution feature can be extracted to generate a seal pair structure synergistic anomaly identification map. This map integrates multi-source information from the structure and dynamic levels to represent the coupling degree of structural integrity degradation and mechanical response instability in the spatial domain. Taking a typical scenario as an example, if a region has a defect probability of 0.85 and a response anomaly index of 0.78, the weight intensity of this region in the synergistic anomaly identification map will be significantly higher than that of other single high-value regions, indicating that this region is a key risk point for potential failure of the sealing function. After the generation of the synergistic anomaly identification map, by analyzing the joint weight relationship of the response anomaly index and the defect probability of each sub-region, a set of seal region evaluation rules for hierarchical evaluation can be established. The joint weight relationship is to form the regional risk intensity score by linear superposition or weighted nonlinear combination of the two indicators, which is used to divide the regional sealing performance level. Specifically, the regions with joint scores exceeding threshold A can be rated as severe failure zones, those between threshold B and A as moderate degradation zones, and those below threshold B as stable sealing zones. Each sub-region is assigned a local performance level label according to this scoring system, representing its comprehensive performance in structural integrity and dynamic stability. For example, a region with both high-frequency response fluctuations and obvious seal surface gaps will have a significantly increased joint score, thus being assigned a "high risk" label.Finally, the performance level labels of all sub-regions are input into the spatial clustering algorithm to analyze the spatial distribution of the aggregation, and the number and intensity of each level label are combined for weighted statistics to generate the evaluation parameters of the overall sealing performance of the butterfly valve. Spatial clustering can be achieved through density clustering or K-means method to identify whether the high-risk area has a concentration, thereby inferring the systematic stability of the entire sealing structure. At the same time, by weighting the area weight and risk coefficient of each level label, a comprehensive performance score value is formed, which represents the sealing reliability level of the butterfly valve under the current operating state. The output sealing performance detection result of the butterfly valve will include the total performance score, the main risk area label map, and the local evaluation suggestions for subsequent operation and maintenance or structure repair, which comprehensively supports the intelligent diagnosis and accurate evaluation of complex metal sealing structures.

[0044] In some embodiments, an infrared thermal imaging sequence of the surface thermal distribution response in the sealing pair structure contact process is acquired, and a multi-frame thermal stress evolution atlas of the sealing area at the corresponding time node is constructed; the spatial coding sensing of the electromagnetic field disturbance of the sealing ring in the opening and closing process is performed by the magnetic flux disturbance sensing array, and the time domain response mode of the edge eddy current disturbance field of the sealing structure is extracted; the thermal stress evolution atlas and the time domain response mode of the edge eddy current disturbance field are aligned in space and time domain, and a consistency index matrix of the dynamic coupling of the sealing structure is generated through synchronous disturbance domain similarity clustering analysis; the dynamic deformation uneven area of the corresponding sealing structure is determined according to the time sequence evolution trend and the spatial dispersion degree of the abnormal offset section in the consistency index matrix, and a sealing performance structure response inconsistent area atlas of the butterfly valve is generated.

[0045] Specifically, in the present embodiment, infrared thermography is a non-contact temperature measurement technique that utilizes the differences in surface temperature radiation energy of objects. It can record the energy transfer process of materials in a thermally coupled state with high temporal resolution. During the opening and closing of the butterfly valve, due to local friction, elastic recovery, or deformation of the micro contact interface in the compression or separation state of the sealing surface, uneven heat accumulation occurs between the sealing ring and the valve seat. By continuously collecting infrared thermal image frames, a thermal response evolution sequence containing multiple time nodes can be constructed, and its thermal stress distribution map in the spatial coordinate system can be extracted, thereby generating a multi-frame thermal stress evolution atlas reflecting the dynamic change trend of the structure's thermal force. This atlas has key value in capturing potential heterogeneous characteristics in the sealing contact process, such as local overheating, thermal response delay, or thermal diffusion abnormalities. Subsequently, by deploying a magnetic flux disturbance sensing array on the edge region of the sealing ring structure, the electromagnetic field disturbance of the sealing structure during the entire opening and closing process of the butterfly valve is continuously monitored. The magnetic flux disturbance sensing array is a multi-point magnetic field detection system based on electromagnetic induction principles. It can sense the change in magnetic flux induced by the movement, stress change, or deformation of metal structures, especially in the edge region where eddy currents are prone to occur. During the opening and closing of the butterfly plate, the edge of the sealing ring will cause temporal fluctuations in local electromagnetic disturbance signals due to small changes in material magnetic permeability, structure thickness, or contact angle. By extracting the time-series magnetic flux density change data recorded by each sensing node and performing encoding compression and filtering denoising, the time-domain response mode of the sealing structure edge eddy current interference field can be constructed to reflect the coupling relationship between electromagnetic response and deformation process at the structure boundary. For example, when there is a gap change or contact relaxation in a certain sealing section, the corresponding magnetic communication signal may exhibit enhanced short-period fluctuations or abnormal amplitude rises. After obtaining the thermal stress evolution atlas and the time-domain response mode of the edge eddy current interference field, the two types of data are aligned in space and time domain by constructing a time synchronization reference axis and performing spatial coordinate mapping. This process involves registration operations based on image timestamps and sensing point spatial positions to ensure that the infrared thermal image frames and magnetic flux disturbance data have consistent data mapping relationships in the same spatial reference system and time nodes. After registration, a synchronous disturbance domain similarity clustering analysis method can be applied to group regions based on the similarity of their dynamic evolution patterns in thermal response and magnetic flux disturbance. Similarity clustering analysis is an unsupervised multivariate classification method that divides regions with highly correlated change trends into the same category by calculating the similarity distance of data in the feature dimension, thereby constructing a consistency index matrix of the dynamic coupling of the sealing structure. This matrix represents the coupling stability degree between thermal response and magnetic disturbance in a spatial region, for example, a region with a matrix value close to 1 represents complete consistency in thermal stress evolution and electromagnetic disturbance mode, and a value close to 0 represents a large response deviation.Finally, the time sequence variation trend of the low coupling value area in the consistency index matrix and the dispersion degree in the spatial domain are analyzed, and the area of the sealing structure with inconsistent response in the dynamic opening and closing process can be identified. The abnormal offset section refers to the spatial area that presents mutation, severe fluctuation or continuous decline on the coupling index curve, and the spatial dispersion degree reflects the distribution density of these abnormal areas in the structure. By jointly judging these two characteristics, a sealing performance inconsistent area map of the structure can be generated, which identifies the key risk sections that may cause the sealing performance to deteriorate due to local instability of the material, deformation hysteresis or structural coupling damage, etc. For example, the long strip-shaped low consistency area in the map may indicate that there is a certain intermittent contact zone on the annular sealing surface, which provides accurate criteria for subsequent structural compensation design and local repair.

[0046] According to the optical detection-based eccentric butterfly valve performance detection method provided in the application, the light transmission intensity distribution of the contact area of the sealing pair and the interference structure light reflection image are jointly acquired, the contact pressure distribution matrix is constructed, and the floating response trajectory in the opening and closing process is combined for spatial comparison, so that the leakage warning area and the corresponding dynamic abnormal features are identified in advance, and the safety and operation stability of the system are improved.

[0047] Figure 2 The optical detection-based eccentric butterfly valve performance detection system provided in the embodiment of the application can be used to implement the optical detection-based eccentric butterfly valve performance detection method in the foregoing embodiment. As shown in the optical detection-based eccentric butterfly valve performance detection system, the optical detection-based eccentric butterfly valve performance detection system mainly includes: Figure 2

[0048] The acquisition module 10 is configured to acquire a first image of the light transmission intensity distribution of the contact area between the sealing pair structure of the bidirectional full-metal hard sealing four-eccentric butterfly valve.

[0049] The analysis module 20 is configured to acquire the reflection interference image of the structured light on the sealing pair structure according to the phase modulation structure grating light field, and jointly analyze the first image to generate a contact pressure distribution matrix.

[0050] The construction module 30 is configured to extract corresponding multi-dimensional image features based on the first image and the contact pressure distribution matrix to generate a leakage warning area map.

[0051] The determination module 40 is configured to spatially compare the leakage warning area map with the floating response trajectory of the sealing pair structure in the opening and closing process to determine the corresponding floating response features of the leakage warning area in the opening and closing process.

[0052] The generation module 50 is configured to generate a butterfly valve sealing performance detection result according to the abnormal degree of the floating response features. ​

[0053] In an optional implementation of the embodiment, the analysis module is specifically configured to: acquire the light transmission intensity spatial distribution map of the sealed pair structure in the closed state according to the first image, and establish a spatial reference coordinate system of the initial sealed pair structure; generate a grating light field with preset phase modulation parameters by controlling the structured light projection unit, project the grating light field to the sealed pair structure according to the spatial reference coordinate system, and collect a reflection interference image sequence formed by the structured light on the sealed pair structure; generate a composite interference profile of the sealing contact interface by performing fringe density distribution extraction and phase reconstruction on the reflection interference image sequence, and combining the light transmission intensity variation characteristics of the same spatial region in the first image; and generate a contact pressure distribution matrix according to the spatial difference mapping relationship between the phase variation gradient of the interference fringes and the transmission intensity in the composite interference profile.

[0054] In an optional implementation of the embodiment, the construction module is specifically configured to: perform spatial registration on the pixel coordinate relationship of the corresponding regions of the first image and the contact pressure distribution matrix, and generate a composite image feature set with unified spatial indexes; extract a multi-dimensional structural feature vector representing the sealing abnormal behavior by jointly counting the composite sealing features of each sub-region in the composite image feature set; input the multi-dimensional structural feature vector into a preset defect label classification model, perform defect label classification on each sub-region in the composite image feature set, determine a corresponding defect label classification result; generate a sealed pair structure defect distribution map according to the defect label classification result, and mark the regions with a defect level higher than a preset threshold in the sealed pair structure defect distribution map as a leakage warning region to generate a leakage warning region map.

[0055] In an optional implementation of the embodiment, the generation module is further configured to: control the opening and closing of the sealed pair structure according to a preset opening and closing time period, and collect an image sequence of a preset optical speckle pattern of the sealed pair structure in consecutive time frames during the opening and closing process; extract the displacement of the optical speckle pattern in the spatial coordinate system by performing sub-pixel registration on the optical speckle patterns between adjacent time frames in the image sequence, and generate a speckle displacement vector sequence; construct a sealing ring surface displacement field corresponding to each time frame by performing time dimension slicing analysis on the speckle displacement vector sequence, and determine the local dynamic response feature map of each time frame according to the spatial displacement continuity and amplitude distribution characteristics; track and superimpose the high response trajectories of each time frame according to the spatiotemporal continuous distribution of the displacement variation abnormal regions in the local dynamic response feature map, and generate a floating response trajectory of the baffle sealing ring in the complete opening and closing period.

[0056] In an optional implementation of the embodiment, the determining module is specifically configured to: establish a two-dimensional space mapping relationship of the sealing pair structure according to the spatial coordinate information of each defect region in the leakage early warning region map and the space-time position sequence of each response path in the floating response trajectory, and construct a comparison reference coordinate system; extract a displacement sub-region corresponding to the early warning region in the comparison reference coordinate system by performing spatial segmentation on the displacement field of each time frame in the floating response trajectory, and generate a local response trajectory group matched with each leakage early warning region; extract a dynamic response feature vector corresponding to the leakage early warning region by jointly analyzing the response directionality, displacement amplitude fluctuation rate, and response disturbance mode of the continuous time frames in the local response trajectory group; and determine the floating response features of each leakage early warning region in the opening and closing process according to the time sequence evolution law and spatial aggregation degree of each feature parameter in the dynamic response feature vector.

[0057] In an optional implementation of the embodiment, the generating module is specifically configured to: generate a multi-dimensional response anomaly index distribution map according to the dynamic response anomaly feature parameters of each response region in the floating response feature; extract the joint distribution features of the response anomaly index and the defect probability in each spatial sub-region by performing spatial superposition analysis on the multi-dimensional response anomaly index distribution map and the defect probability distribution map output by the defect label classification model, and generate a sealing pair structure cooperative anomaly identification map; establish a sealing region hierarchical evaluation rule according to the joint weight relationship of the response anomaly index and the defect probability in the sealing pair structure cooperative anomaly identification map, and divide each sub-region into a sealing performance level to determine a corresponding local performance level label; obtain a butterfly valve overall sealing performance evaluation parameter by performing spatial clustering and level weighted statistics on all local performance level labels, and output a butterfly valve sealing performance detection result.

[0058] In an optional implementation of the embodiment, the generating module is further configured to: obtain an infrared thermal imaging sequence of the surface thermal distribution response in the contact process of the sealing pair structure, and construct a multi-frame thermal stress evolution map of the sealing region at a corresponding time node; extract the time domain response mode of the edge eddy current interference field of the sealing structure by performing spatial coding sensing on the electromagnetic field disturbance of the sealing ring in the opening and closing process through the magnetic flux disturbance sensing array; perform space-time domain alignment on the thermal stress evolution map and the time domain response mode of the edge eddy current interference field, and generate a consistency index matrix of the dynamic coupling of the sealing structure by performing synchronous disturbance domain similarity clustering analysis; determine the dynamic deformation uneven region of the corresponding sealing structure according to the time sequence evolution trend and spatial dispersion degree of the abnormal offset section in the consistency index matrix, and generate a butterfly valve sealing performance structure response inconsistent region map.

[0059] According to the scheme provided by the application, the optical detection-based eccentric butterfly valve performance detection system is provided, the light transmission intensity distribution and the interference structure light reflection image of the sealing pair contact area are jointly acquired, the contact pressure distribution matrix is constructed, and the spatial comparison is performed in combination with the floating response track in the opening and closing process, so that the leakage early warning area and the corresponding dynamic abnormal features are identified in advance, and the safety and operation stability of the system are improved.

[0060] According to the scheme provided by the application, Figure 3 An electronic device is provided for the embodiments of the application. The electronic device can be used to implement the optical detection-based eccentric butterfly valve performance detection method in the foregoing embodiments, and mainly includes:

[0061] The memory 301, the processor 302, and the computer program 303 stored in the memory 301 and executable on the processor 302 are communicatively connected. When the processor 302 executes the computer program 303, the optical detection-based eccentric butterfly valve performance detection method in the foregoing embodiments is implemented. The number of processors can be one or more.

[0062] The memory 301 can be a high-speed random access memory (RAM) or a non-volatile memory such as a disk memory. The memory 301 is used to store executable program codes, and the processor 302 is coupled with the memory 301.

[0063] Further, the embodiments of the application also provide a computer readable storage medium, which can be arranged in the electronic device in the above embodiments. The computer readable storage medium can be the memory in the above embodiments. Figure 3

[0064] The computer readable storage medium stores a computer program, and the program is executed by the processor to implement the optical detection-based eccentric butterfly valve performance detection method in the foregoing embodiments. Further, the computer readable storage medium can also be a U disk, a mobile hard disk, a read-only memory (ROM), a RAM, a magnetic disk or an optical disk, and various media that can store program codes.

[0065] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the system, the system and the unit described above can refer to the corresponding process in the foregoing method embodiments, which will not be described here.

[0066] ​The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application or the entire or part of the technical solutions that essentially contribute to the prior art can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0067] The above-described embodiments are merely used to illustrate the technical solutions of the present application, rather than limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for some of the technical features; and these modifications or replacements do not cause the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for testing the performance of an eccentric butterfly valve based on optical detection, characterized in that, include: Obtain the first image of the light transmission intensity distribution of the sealing pair structure of the bidirectional all-metal hard-seal four-eccentric butterfly valve; The reflection interference image of structured light on the sealing substructure is acquired based on the phase-modulated structured grating light field, and combined with the first image for joint analysis to generate a contact pressure distribution matrix; Based on the first image and the contact pressure distribution matrix, the corresponding multi-dimensional image features are extracted to generate a leak warning area map. The leakage warning area map is spatially compared with the floating response trajectory of the sealing pair structure during the opening and closing process to determine the floating response characteristics of the leakage warning area during the opening and closing process. The butterfly valve sealing performance test results are generated based on the degree of anomaly of the floating response characteristics; The step of acquiring the reflection interference image of structured light on the sealing substructure based on the phase-modulated structured grating light field, and performing joint analysis with the first image to generate a contact pressure distribution matrix includes: Based on the first image, obtain the spatial distribution map of light transmission intensity of the sealing pair structure in the closed state, and establish the corresponding spatial reference coordinate system of the initial sealing pair structure; A grating light field with preset phase modulation parameters is generated by controlling the structured light projection unit, and the grating light field is projected onto the sealing substructure according to the spatial reference coordinate system. The reflection interference image sequence formed by the structured light on the sealing substructure is then acquired. By extracting the fringe density distribution and reconstructing the phase of the reflected interference image sequence, and combining the light transmission intensity variation characteristics of the same spatial region in the first image, a composite interference profile of the sealed contact interface is generated. Based on the spatial difference mapping relationship between the phase change gradient of the interference fringes and the transmission intensity in the composite interference profile, a contact pressure distribution matrix is ​​generated.

2. The method for testing the performance of an eccentric butterfly valve based on optical detection according to claim 1, characterized in that, The step of extracting corresponding multidimensional image features to generate a leak warning area map based on the first image and the contact pressure distribution matrix includes: Spatial registration is performed on the pixel coordinates of the first image and the corresponding region of the contact pressure distribution matrix to generate a composite image feature set with a unified spatial index. By jointly statistically analyzing the composite sealing features of each sub-region in the composite image feature set, a multi-dimensional structural feature vector characterizing abnormal sealing behavior is extracted. The multidimensional structural feature vector is input into a preset defect labeling classification model to classify defects in each sub-region and determine the corresponding defect labeling classification result. Based on the defect marking classification results, a defect distribution map of the sealing pair structure is generated, and areas in the defect distribution map of the sealing pair structure with defect levels higher than a preset threshold are marked as leakage warning areas, generating a leakage warning area map.

3. The method for testing the performance of an eccentric butterfly valve based on optical detection according to claim 1, characterized in that, Before the step of spatially comparing the leakage warning area map with the floating response trajectory of the sealing pair structure during the opening and closing process to determine the floating response characteristics corresponding to the leakage warning area during the opening and closing process, the method further includes: The opening and closing of the sealing pair structures are controlled according to the preset opening and closing time period, and the image sequence of the preset optical speckle pattern of the sealing pair structure in continuous time frames is acquired during the opening and closing process. By performing sub-pixel registration on the optical speckle patterns between adjacent time frames in the image sequence, the displacement of the optical speckle patterns in the spatial coordinate system is extracted to generate a speckle displacement vector sequence. By performing time-dimensional slice analysis on the speckle displacement vector sequence, the surface displacement field of the sealing ring corresponding to each time frame is constructed, and the local dynamic response feature map of each time frame is determined based on the spatial displacement continuity and amplitude distribution characteristics. Based on the spatiotemporal continuous distribution of the abnormal displacement regions in the local dynamic response feature map, the high response trajectories of each time frame are tracked and superimposed to generate the floating response trajectory of the butterfly plate sealing ring within a complete opening and closing cycle.

4. The method for testing the performance of an eccentric butterfly valve based on optical detection according to claim 3, characterized in that, The step of spatially comparing the leakage warning area map with the floating response trajectory of the sealing pair structure during the opening and closing process to determine the floating response characteristics of the leakage warning area during the opening and closing process includes: Based on the spatial coordinate information of each defect area in the leakage early warning area map and the spatiotemporal position sequence of each response path in the floating response trajectory, a two-dimensional spatial mapping relationship of the sealing pair structure is established, and a comparison reference coordinate system is constructed. By spatially segmenting the displacement field of each time frame in the floating response trajectory, the displacement sub-regions corresponding to the warning areas in the comparison reference coordinate system are extracted, and local response trajectory groups matching each leakage warning area are generated. By jointly analyzing the response directionality, displacement amplitude fluctuation rate, and response disturbance mode of continuous time frames in the local response trajectory group, dynamic response feature vectors corresponding to the leakage warning area are extracted. Based on the temporal evolution and spatial clustering of each feature parameter in the dynamic response feature vector, the floating response characteristics of each leakage warning area during the opening and closing process are determined.

5. The method for testing the performance of an eccentric butterfly valve based on optical detection according to claim 2, characterized in that, The step of generating butterfly valve sealing performance test results based on the degree of anomaly of the floating response characteristics includes: A multidimensional response anomaly index distribution map is generated based on the dynamic response anomaly characteristic parameters of each response region in the floating response characteristics. By performing spatial overlay analysis on the multidimensional response anomaly index distribution map and the defect probability distribution map output by the defect labeling classification model, the joint distribution features of response anomaly index and defect probability in each spatial sub-region are extracted to generate a collaborative anomaly identification map of the sealing pair structure. Based on the joint weighting relationship between the response anomaly index and the defect probability in the collaborative anomaly identification diagram of the sealing pair structure, a hierarchical evaluation rule for the sealing area is established, and the sealing performance level of each sub-region is divided to determine the corresponding local performance level label. By performing spatial clustering and weighted statistical analysis on all local performance level labels, the overall sealing performance evaluation parameters of the butterfly valve are obtained, and the sealing performance test results of the butterfly valve are output.

6. The method for testing the performance of an eccentric butterfly valve based on optical detection according to claim 1, characterized in that, The method further includes: Infrared thermal imaging sequences of the surface thermal distribution response during the contact process of the sealing pair structure were obtained, and multi-frame thermal stress evolution maps of the sealing area at corresponding time nodes were constructed. The electromagnetic field disturbance of the sealing ring during the entire opening and closing process is spatially encoded and sensed by a magnetic flux disturbance sensing array, and the time-domain response mode of the eddy current interference field at the edge of the sealing structure is extracted. The thermal stress evolution spectrum and the temporal response mode of the edge eddy current disturbance field are aligned in the spatial and temporal domains, and a consistency index matrix of dynamic coupling of the sealed structure is generated through synchronous disturbance domain similarity clustering analysis. Based on the temporal evolution trend and spatial dispersion of the abnormal offset segments in the consistency index matrix, the dynamic deformation uneven region of the corresponding sealing structure is determined, and a region map of inconsistent structural response of butterfly valve sealing performance is generated.

7. A performance testing system for eccentric butterfly valves based on optical detection, characterized in that, The optical detection-based eccentric butterfly valve performance testing system is used to implement the optical detection-based eccentric butterfly valve performance testing method of claim 1, wherein the optical detection-based eccentric butterfly valve performance testing system comprises: The acquisition module is used to acquire a first image of the light transmission intensity distribution in the contact area between the sealing pairs of a bidirectional all-metal hard-seal four-eccentric butterfly valve. The analysis module is used to acquire the reflection interference image of structured light on the sealing substructure based on the phase-modulated structured grating light field, and perform joint analysis with the first image to generate a contact pressure distribution matrix; The construction module is used to extract corresponding multi-dimensional image features and generate a leak warning area map based on the first image and the contact pressure distribution matrix. The determination module is used to spatially compare the leakage warning area map with the floating response trajectory of the sealing pair structure during the opening and closing process, and determine the floating response characteristics of the leakage warning area during the opening and closing process. The generation module is used to generate butterfly valve sealing performance test results based on the degree of abnormality of the floating response characteristics.

8. An electronic device, characterized in that, Includes memory and processor, of which: The processor is used to execute computer programs stored in the memory; When the processor executes the computer program, it implements the steps in the optical detection-based eccentric butterfly valve performance testing method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps in the optical detection-based eccentric butterfly valve performance testing method according to any one of claims 1 to 6.

Images