147results about How to "Improve detection stability" patented technology

The invention relates to an Internet of Vehicles multi-agent cooperative sensing method and system. The method comprises the following steps: constructing a collaborative sensing network, wherein the collaborative sensing network comprises a self-agent and a plurality of collaborative agents; acquiring and processing sensing data through the collaborative sensing network; performing feature extraction to obtain intermediate features; self-adaptive sparsification is carried out to obtain sparse features, and the sparse features are compressed and transmitted to a self-agent; performing time sequence feature enhancement on the features of all the agents at the self-agent end; fusing the features to obtain fused features; and constructing an Internet of Vehicles perception model, and realizing perception by the detection model according to the fused features. According to the method, the calculation complexity of traditional global attention is reduced from the square level to the linear level through an adaptive sparsification mechanism, the calculation overhead is remarkably reduced while the multi-agent feature interaction precision is kept, and the method is more suitable for real-time operation on the vehicle-mounted edge equipment with limited resources.

The invention relates to the technical field of visual inspection, in particular to a stamping part defect detection method based on vision, which comprises the following steps: S1, shooting setting; s2, synchronizing video streams; s3, acquiring a part bounding box and a hole position bounding box; s4, associating part bounding boxes belonging to the same stamping part among the frame part images, and associating hole position bounding boxes belonging to the same screw hole position; s5, correspondingly associating the bounding boxes of the target parts belonging to the same stamping part in the part image; s6, when a hole position boundary frame matched with the view angle template image of the corresponding view angle exists in any target detection image under any view angle, determining that a screw hole position exists; and when the matched hole position boundary frame is not found in the target detection image under the corresponding visual angle in the visual angle template images of all the visual angles, the screw hole position is missing. The stamping part defect detection device can be used for defect detection of stamping parts in the conveying process, the consistency of detection results is improved, and the detection efficiency is improved.

The invention discloses a steel structural member local corrosion intelligent identification and evaluation method based on deep learning, and the method comprises the steps: obtaining a steel structure surface image through an imaging device, carrying out the feature extraction and corrosion region identification of an input image through a YOLOv12 semantic segmentation model which introduces a CBAM self-attention mechanism, and generating a pixel-level corrosion mask; the actual physical area of the corrosion area is automatically calculated based on an area calculation algorithm of conversion of mask integration and ground sampling distance; the residual bearing performance of the corroded component is evaluated in combination with the corrosion area ratio and the morphological characteristic prediction model; and finally realizing visual display and report output of detection results, area data and performance indexes through a graphical interface. According to the invention, full-process automatic processing of local corrosion of the steel structural member from detection, segmentation, quantitative measurement and calculation to member performance evaluation is realized, and closed-loop intelligent evaluation from visual detection to mechanical evaluation is realized.

The invention discloses an AI substitute inspection method and system based on multi-modal perception, and relates to the technical field of intelligent detection. According to the method, images, audios, environmental parameters and equipment state data of an inspection area are synchronously collected through a multi-modal sensor array, and after space-time alignment preprocessing, multi-modal features are extracted and input into a pre-trained fusion model to generate unified feature representation. The fusion model adopts a Transform architecture to realize cross-modal feature interaction, an anomaly detection algorithm is used to identify and classify anomalies, a confidence coefficient is calculated in combination with an evidence theory, and finally an inspection report containing anomaly positions, types and processing suggestions is generated. According to the invention, the information limitation of single-mode inspection is solved, the detection accuracy and robustness in a complex environment are improved, unmanned inspection in a high-risk scene is realized, the labor cost is reduced, the safety is guaranteed, and the method is suitable for automatic inspection in the fields of industrial equipment, public facilities and the like.

The invention relates to the technical field of roadbed detection, and discloses a roadbed compaction quality detection method based on ultrasonic wave collaboration, and the roadbed compaction quality detection method comprises the following steps: collecting full-wave band ultrasonic echo signals, and obtaining current environment data; calculating a longitudinal wave speed, a transverse wave speed and a wave speed ratio, and extracting an attenuation coefficient and a frequency dispersion characteristic of the ultrasonic echo signal to obtain an acoustic characteristic set; inputting the acoustic feature set and the environmental data into an environmental compensation model, outputting a standard acoustic parameter set subjected to temperature, humidity and water content correction, analyzing the standard acoustic parameter set through a soil texture classification algorithm, and judging the current soil texture type; and combining the standard acoustic parameter set and the current soil texture type into a multi-dimensional feature vector, inputting the multi-dimensional feature vector into a compactness calculation model, and calculating to obtain a roadbed compactness value of the current measuring point. According to the method, high-precision compaction degree numerical calculation is realized, and intelligent and visual whole-process support is provided for roadbed construction quality.

The invention discloses a method for determining the content of rubber in a rubber-containing plant based on gel permeation chromatography, and relates to the technical field of analytical chemistry, and the method specifically comprises the following steps: S1, sample pretreatment: drying a rubber-containing plant sample and grinding into powder; s2, rubber extraction: performing ultrasonic extraction on the powder by using a mobile phase of gel permeation chromatography as an extraction solvent, centrifuging, and collecting supernate to obtain an extracting solution; s3, GPC (Gel Permeation Chromatography) detection: filtering the extracting solution, and injecting the filtered extracting solution into a gel permeation chromatographic instrument with a single chromatographic column for detection. The complete plant rubber content determination method is constructed by combining a total rejection GPC principle with an efficient sample pretreatment process, the defects of tedious pretreatment, low detection efficiency or inaccurate detection in the prior art are overcome, the method has the advantages of rapidness, accuracy and the like, extraction and detection parameters can be flexibly adjusted according to different plant samples, and the method is suitable for large-scale popularization and application. The method is suitable for quality analysis and breeding research of various gelatin-containing plants, so that the compatibility is improved.

The application belongs to the technical field of marine chemical analysis, and relates to an underwater in-situ mass spectrometer acquisition control system and method. The method comprises the following steps: a vacuum extraction step: the coarse extraction mode and the fine extraction mode are adaptively switched, the ion detection device is vacuum-extracted until the pressure is reduced to a preset threshold, and a required vacuum degree is maintained; a preheating preparation step: the membrane sampling device is initially sampled, and heating is performed at the same time until a preset working temperature is reached and the temperature is maintained constant; a dissolved gas formal detection step: the membrane sampling device is continuously supplied with water samples to be detected, and in-situ mass spectrum detection of dissolved gas in the water samples is completed; an instrument cabin gas discharge step: when it is detected that the underwater in-situ mass spectrometer floats to the water surface and the air pressure in the instrument cabin exceeds an upper limit threshold of gas discharge, the instrument cabin is discharged; a liquid leakage protection step: when liquid leakage occurs in the sampling cabin and the two core contacts are short-circuited, the high-pressure electromagnetic valve in the normally open state is closed to quickly block the liquid from entering the ion detection device.

This invention discloses a crankshaft axis inspection instrument based on machine vision, belonging to the field of crankshaft inspection technology. The machine vision-based crankshaft axis inspection instrument includes an inspection table and a primary lead screw slide mounted on the inspection table. A primary slide is slidably mounted on the primary lead screw slide, and a mounting frame is fixed on the primary slide. A vision camera for visually inspecting the crankshaft axis is mounted on the mounting frame. A support frame is vertically fixed on the mounting frame, and an illumination component coaxial with the vision camera is mounted on the support frame. Two sets of follow-up support wheels are provided below the crankshaft main journal to prevent deformation due to their own weight. A base component is provided on the support frame to adjust the position and angle of the follow-up support wheels. A base is mounted on the base component, and an adaptive support sensing component is provided on the base to sense the pressure of the follow-up support wheels supported below the crankshaft main journal. This inspection instrument can support and limit the crankshaft main journal, reduce axis inspection interference, and improve the accuracy of visual inspection.

The application discloses a detection method and device for a charging pile and a charging pile system, and belongs to the field of charging piles. The detection method for the charging pile comprises the following steps: acquiring radar data collected by a radar module; based on the radar data, a detection strategy corresponding to a detection target is used for detection to obtain a detection result; the detection target comprises area occupancy detection of a parking area and / or cable state detection. The detection method for the charging pile has high detection accuracy and stability, can predict whether the cable has a risk of being stolen, has a certain advancement, and is helpful to improving the use experience of users.

The invention relates to a nano antibody for specifically recognizing a tacrolimus-antibody compound and application of the nano antibody, and belongs to the technical field of biological detection. A variable region of the nano antibody has three complementary determining regions CDR1, CDR2 and CDR3, can specifically recognize a tacrolimus-antibody compound, and can be used for detecting drugs by a double-antibody sandwich method; the kit is widely suitable for various immunodetection technologies such as an enzyme-linked immunosorbent assay (ELISA), a chemiluminescence immunoassay (CLIA), an electrochemical luminescence immunoassay (ECLIA), a fluorescence immunoassay (FIA), an immunochromatography test strip method and the like. According to the method, the limitation of a traditional competition method is overcome, high-sensitivity, high-specificity and high-stability detection is realized, the method can be adapted to various detection platforms, and a portable pretreatment process without centrifugation can also be realized.

The present application relates to the technical field of industrial vision detection, and discloses a green color printing surface defect detection method based on computer vision. The method comprises the following steps: collecting real-time image flow and performing multi-scale decomposition to generate a feature vector sequence; constructing a dynamic defect template set by using a generative adversarial network; encoding features by using a convolutional neural network, calculating a similarity matrix with the template set, and generating a defect matching index; adjusting the detection threshold value adaptively according to the defect matching index, and obtaining a defect probability distribution map through clustering analysis; identifying a defect aggregation area through space-time domain analysis, and starting a local re-inspection mechanism; updating a defect feature prototype library and optimizing network weights by using re-inspection data, and forming a self-learning closed loop; and finally performing full-process quality consistency inspection and outputting a report. The present application realizes dynamic adaptation and continuous optimization of the detection process.

The utility model relates to the technical field of coating die precision control, specifically relates to an actuating mechanism and coating machine. An actuating mechanism, include: integrated seat body, displacement piece, be located integrated seat body one side, displacement sensor, displacement sensor be located integrated seat body, displacement sensor is equipped with elastic probe, elastic probe with displacement piece is connected, elastic probe with displacement piece same side setting, displacement piece is driven ground to move towards or away from integrated seat body, displacement probe follows displacement piece movement, displacement sensor senses displacement piece's movement distance. The utility model provides an actuating mechanism and coating machine to solve in traditional actuating mechanism, multiple installation error's superposition leads to the overall assembly error of screw rod and displacement sensor between big, influence the whole mechanism's assembly precision problem.

The invention discloses a high-precision double-side thickness synchronous detection system for a notebook computer shell, and relates to the field of appearance detection, and the system comprises a positioning module which is used for receiving a to-be-detected notebook computer shell, adsorbing the surface of the shell through a plurality of groups of adjustable vacuum adsorption parts, and carrying out the automatic centering and posture calibration of the shell based on the feedback of a displacement sensor; the detection module integrates symmetrically-arranged detection units and is used for synchronously applying detection signals to preset detection points on the two sides of the positioned shell and collecting original detection data related to the thickness of the two sides; the adsorption pressure is dynamically adjusted along with the surface curvature, the posture is accurately calibrated in a linkage mode, it is ensured that the detection object is stably positioned, the posture deviation is extremely small, the detection signal counteracts the distance and medium attenuation influence through amplitude real-time compensation, the original data quality is optimized in cooperation with self-adaptive filtering denoising and accurate amplitude calibration, and the detection accuracy is improved. And temperature, humidity and other environmental factors and material characteristic correction are fused, so that the thickness calculation result is closer to reality.

The application relates to a detection system for an SDN terminal device, wherein a system upgrade interface, an optical port, an electric port and a controller interface are arranged on the SDN terminal device; the detection system comprises a detection device and a remote server; the detection device comprises a shell and a system upgrade assembly, an optical port detection assembly, an electric port detection assembly, an SDN simulation control assembly and a main controller arranged in the shell; the system upgrade assembly is connected with the system upgrade interface; the optical port detection assembly is connected with the optical port; the electric port detection assembly is connected with the electric port; and the SDN simulation control assembly is connected with the controller interface. Compared with the prior art, the application can detect the electric port, the optical port and the voice port of the SDN terminal device and perform system updating operation under the condition that the SDN terminal device is disconnected with the SDN controller, and the detection function is perfect and the detection effect is good.

The food freshness detection device specifically comprises a main body, and a sampling assembly, a detection assembly and an air exhaust assembly which are sequentially arranged on the main body, the sampling assembly comprises a shell, a temperature adjusting part and a triggering part, the shell is arranged on the main body, and a sampling cavity is concavely formed in the end, close to food, of the shell; the temperature adjusting part is arranged on the sampling cavity; the detection assembly comprises a detection cavity and a detector, the detection cavity is communicated with the sampling cavity, a gas outlet is formed in the side, away from the sampling cavity, of the detection cavity, and the gas extraction assembly is used for sucking gas in the sampling cavity into the detection cavity and discharging the gas from the gas outlet; the detector is arranged in the detection cavity and is used for detecting the concentration of ammonia gas in the detection cavity. By adjusting the food temperature, triggering air exhaust detection at a proper temperature and timely discharging residual gas, the interference of low temperature and gas residue on ammonia gas detection is effectively avoided, and the accuracy, stability and reliability of food freshness detection are improved.

The present application relates to a kind of vehicle speed detection methods based on outer pole face image, belong to the field of automatic driving.The method includes: S1: periodically capture the image of the road in front of vehicle, obtain time-space distance image;S2: the curve cluster surface of the time-space distance image obtained is divided into several parts;S3: the curve corresponding to each part obtained;S4: by its analytic differential, the speed curve of each part is obtained;S5: the generation of full interval speed curve, according to the displacement of feature point, the speed of vehicle is calculated.The present application is based on the principle of visual kinematics, using the image obtained by vehicle-mounted camera and distance sensor, the speed of vehicle is calculated by observing the motion of feature point in image.Compared with the existing speed detection method based on physical sensor, the method is lower in cost, but the principle is simple, and the accuracy is also higher.The method can be used alone, can also be combined with the data of other sensors, improve detection accuracy and stability.

The invention discloses a visual detection method and device, computer equipment and a storage medium. The image quality is effectively improved through the preprocessing steps of illumination compensation, denoising, background suppression and the like. The local gray scale difference of the potential target area is further amplified through contrast enhancement; a pre-trained convolutional neural network is introduced, the convolutional neural network adopts a YOLOv5 network structure, and multi-scale feature extraction, residual enhancement and feature fusion are combined, so that the system can capture target visual features of different sizes and different depth levels at the same time. The extracted multi-level effective feature vectors are used for training a classification model, and the system can gradually form the target recognition capability for a specific machining workpiece. In a real detection stage, the model quickly deduces the to-be-detected image, and can maintain high detection precision and stability in a complex production environment, thereby meeting the requirements of efficient, real-time and high-reliability industrial visual detection.

This invention relates to a fluid velocity and flow rate detection device, and pertains to the field of fluid detection technology. The device includes: a fixed small steel plate and an elastic sensing unit; both the fixed small steel plate and the elastic sensing unit are disposed inside the fluid pipe to be detected; wherein, the lower end of the fixed small steel plate is fixedly connected to the inner wall of the fluid pipe, and a fixing part is provided at its upper end; the elastic sensing unit includes at least two elastic steel plates, one end of each elastic steel plate converging and connecting to the fixing part, and the elastic steel plates are radially distributed with the fixing part as the center; a first sensor is assembled at the other end of each elastic steel plate. This invention has a wide range of applicable media, strong anti-interference ability, and stable detection.

The invention belongs to the technical field of optical thin film detection, and particularly discloses an optimization method of an optical thin film detection system, which is used for solving the problems of weak interference resistance, poor pollution resistance and insufficient complex scene adaptation of the traditional detection system. The method comprises the following steps: pre-selecting a light needle beam with a needle-point-shaped cross section as a light source, defining a source plane light field, and constructing a limited energy generation condition through apodization operation; setting a plurality of groups of regulation and control parameters for three transmission states of diffraction, non-diffraction and focusing; building an experiment system to generate a directional light needle light beam, simulating pollution and interference scenes, and collecting light field data of different transmission nodes before and after interference; and quantizing the light field similarity through a witness function, screening an optimal regulation and control parameter, and integrating the optimal regulation and control parameter to a detection system. According to the invention, the anti-interference and information stability are improved from the light source essence, the method is adaptive to ultrathin, multi-layer thin film and complex environment detection, the detection precision and long-term stability are obviously improved, and both practicability and economy are taken into account.

The application relates to the technical field of flaw detection, and discloses a flaw detection device for numerical control machine tool part machining, which comprises a conveying frame with supporting legs, a conveying mechanism is arranged in the conveying frame, the upper surface of one end of the conveying frame is fixedly connected with a detection box, entrances and exits are arranged on the two sides of the detection box close to the chain plate, the middle part of the upper surface of the detection box is fixedly connected with a mounting rod, the lower end of the mounting rod penetrates into the inside of the detection box and is fixedly connected with a mounting block, the bottom surface of the mounting block is fixedly connected with a visual detection camera, one side of the detection box close to the conveying frame is connected with a moving mechanism, and the output end of the moving mechanism is connected with an L-shaped plate. The flaw detection device for numerical control machine tool part machining is provided with the L-shaped plate carrying a standard part, can automatically move to the position below the camera in the detection gap to complete self-checking and calibration, does not need to stop and maintain, and solves the problem of efficiency loss caused by traditional regular stop and calibration.

The application discloses a multi-degree-of-freedom adjusting probe array for battery module equalization detection, and relates to the technical field of battery detection.The multi-degree-of-freedom adjusting probe array comprises a base, a plurality of mounting columns are symmetrically and fixedly installed on the base, an electric telescopic rod is fixedly installed in each mounting column, a supporting frame is fixedly and jointly installed at the telescopic ends of the plurality of electric telescopic rods, a plurality of supporting frames are equidistantly clamped in the supporting frame, and a plurality of fixing cylinders are slidably installed in each supporting frame.The application solves the problem that the existing battery module accumulates position errors in the assembly link, causes small height differences in the height direction of each pole in the same module, and causes the surfaces of some battery poles to present a small inclination after welding, which leads to unstable contact between the probe and the pole, and the pole spacing, pole height and arrangement mode of battery modules of different capacities and specifications are greatly different, the common probe array cannot be flexibly adjusted, and the production line efficiency is seriously affected.

The utility model discloses a bottle feeding speed adjusting system on a glass bottle detection production line, which relates to the technical field of glass bottle production, and comprises a main body structure and a frame body, the top of the frame body is fixedly connected with a conveying bottom belt, one side of the frame body is fixedly connected with a bottle conveying crawler belt, one side of the bottle conveying crawler belt is provided with a speed changing part, and the speed changing part is provided with a rotating shaft. The two sides of the top of the conveying bottom belt are each provided with a synchronous belt, one side of each synchronous belt is provided with a bottle removing piece, and the two sides of the top of the frame body are each fixedly connected with a side plate and an adjusting assembly; automatic adjustment of the distance between the synchronous belts is achieved through the adjusting assembly, manpower input is greatly reduced, the adjusting time is greatly shortened, the production efficiency is remarkably improved, the equipment maintenance cost is reduced, dynamic adjustment of the glass bottle conveying speed and automatic removal of inverted bottles are achieved through the design of the speed changing part and the bottle removing part, and the production efficiency is improved. The bottles entering the detection equipment are ensured to be reasonable in spacing and normal in posture, and the detection precision and the equipment operation stability are effectively improved.

The invention discloses a complex weather small target detection system based on multi-scale image enhancement and dynamic domain adaptation, and the system comprises an LMIENet image enhancement module, an SC2f deformable convolution feature extraction module, and a DDA dynamic domain adaptation module, and the detection method of the detection system comprises the following steps: a, collecting a plurality of types of weather scene images, and constructing an original image data set; b, data cleaning and refined labeling; c, performing image enhancement preprocessing; d, dividing a structured data set; e, training and calibrating an LMIENet image enhancement module; f, constructing a detection network integrated with deformable convolution; g, introducing joint training adapted to a dynamic domain; h, pipeline construction and end-to-end optimization; and i, system performance verification. According to the method, multi-scale feature fusion, complex weather preprocessing and dynamic domain self-adaption are integrated, and the detection precision and robustness of the small target under the complex weather condition can be effectively improved.

The application provides a single sample learning path construction method for medical image key point detection. The method comprises: feature extraction and sample similarity measurement on images in a data set to obtain a similarity score; according to the similarity score, a single typical sample representative in a key point area structure is screened out from the data set; a pseudo label generator is trained based on the typical sample; pseudo label information of key points of other samples in the data set is generated by using the pseudo label generator; the pseudo label information is used as a supervision signal to train a key point detector, and a detection network of the key point detector adopts a U-Net network as a backbone network and a pre-trained VGG19 encoder as a feature extractor of the network; and the key point detector is used to detect key points of a target image and predict a key point heat map and a coordinate offset map. The method of the application reduces labeling cost, improves the precision and stability of key point detection, and enhances the generalization ability.

The invention belongs to the crossing field of an automobile active safety system and a high-precision gas sensing technology, and provides a vehicle-mounted alcohol system and method which can avoid interference of other ethanol-containing substances and only detect gas exhaled by a driver. The system defects of an existing vehicle-mounted alcohol detection system in the aspects of reliability, anti-interference performance and integration are overcome. On the basis of traditional alcohol detection and photoacoustic spectrum detection, the detection sensitivity, reliability and accuracy are improved, meanwhile, the device can be integrally applied to a vehicle-mounted alcohol detection system, and the requirement for rapidly and accurately detecting whether a driver drinks alcohol or not in a vehicle in real time is met.

The invention discloses a point cloud data efficient three-dimensional target detection method for an intelligent construction scene, and belongs to the field of engineering management, and the detection method comprises the following specific steps: I, periodically collecting and preprocessing construction site point cloud data, building a continuous time frame sequence, and constructing a construction twinborn model corresponding to a construction site; iI, aligning the semantic topology of the point cloud data and the construction twinborn model, identifying the abnormal area of the construction site, and then analyzing the evolution trajectory of each object on the construction site in real time; according to the method, the occurrence probability of false detection and unreasonable engineering semantics results is remarkably reduced, the detection stability of slowly moving or periodically changing targets is effectively enhanced, the problems of jumping and missing detection which are easily generated by single-frame detection are avoided, the recognition accuracy of shielded components in a complex construction environment is improved, and the construction quality is improved. And the overall detection efficiency is improved while the precision is ensured.

This invention provides a colorimetric / fluorescence dual-mode nanozyme aptamer sensor, its preparation method, and its application in bisphenol A (BPA) detection. The sensor comprises a metal-organic framework nanozyme loaded with a BPA aptamer, hydrogen peroxide, and a chromogenic substrate, TMB. Cu / Zr-MOF oxidizes the colorless TMB to blue oxTMB, producing a UV absorption peak at 652 nm and a fluorescence emission peak at 415 nm. Modification with the BPA aptamer enhances the specific recognition ability of BPA, while simultaneously inhibiting the enzyme activity of Cu / Zr-MOF, leading to weakened colorimetric and fluorescence signals. In the presence of BPA, the aptamer specifically binds to the target and dissociates from the nanozyme surface, restoring enzyme activity and enhancing colorimetric and fluorescence signals. This invention integrates the excellent catalytic ability of nanozymes with the specific recognition ability of aptamers into a single system, achieving highly sensitive BPA detection.

The utility model relates to a kind of visual inspection equipment suitable for pencil, including equipment pedestal, workbench is installed on equipment pedestal, main steering adjusting device is installed in one side of workbench, displacement adjusting device is installed in the other side of workbench, vice steering adjusting device is installed on displacement adjusting device, the position of main steering adjusting device, vice steering adjusting device corresponds, constitutes wire clamping area;Lifting support is installed on equipment pedestal, visual inspection device is installed in the upper end of lifting support, display device is also installed on lifting support.It can realize multifunctional adaptive adjustment capability, and can satisfy all-around dead angle visualization detection.

The application discloses a method, system and device for detecting Tau protein based on machine learning of surface enhanced Raman spectroscopy, and the method comprises the following steps: S1, constructing a SERS signal probe with minimum initial Raman signal; S2, mixing a sample to be detected with the SERS signal probe, and collecting SERS spectra of the mixed system by using a Raman spectrometer; S3, pre-processing the collected SERS spectra, wherein the pre-processing comprises baseline correction, feature peak identification and minimum-maximum normalization processing in sequence, so as to obtain standardized spectral data; and S4, inputting the standardized spectral data into a machine learning regression model which is trained in advance, and automatically outputting a concentration value of Tau protein in the sample to be detected based on the input spectral characteristics by the machine learning regression model. The application can quantitatively detect phosphorylated Tau protein quickly, accurately and with low detection limit under the premise of keeping high specificity, and has high clinical conversion potential.

This invention relates to the field of conveyor belt testing technology, specifically to a tensile strength testing device for conveyor belt production. The device includes a frame, characterized by further comprising a pair of limiting rollers, a pair of movable fixing components, and a fracture protection component. The movable fixing components include a support frame, a guide plate, and a bending roller with rotational power. The two support frames are symmetrically distributed at the top and bottom of the frame. The bottom support frame is fixedly connected to the bottom wall of the frame, while the top support frame is vertically slidable on the frame. The bending roller is rotatably mounted on the support frame. This invention addresses the technical problems of cumbersome clamping, uneven force distribution, and easy tearing at the clamping end in traditional conveyor belt tensile strength testing. By folding the end of the conveyor belt body around the bending roller and simultaneously passing the limiting roller through the arc-shaped structure formed by the bending of the conveyor belt body, friction is used for fixation, achieving efficient and convenient clamping, uniform tensile force distribution, and stable heavy-load testing.