Prestressed tension beam breaking identification early warning system and method

Abstract

The application provides a prestress tensioning bundle breakage identification and early warning system and method, adopts a double closed loop discrimination framework of mechanical slope main judgment combined with video dynamic review, collects force value and elongation data in the steel strand tensioning process, calculates a force value-elongation dynamic slope in real time, preliminarily judges the bundle breakage suspected abnormality according to the slope threshold value and triggers a first level early warning, meanwhile, the review module extracts multi-dimensional features such as inter-frame displacement, bundle body deformation and vibration through video analysis, discriminates false abnormalities caused by construction disturbances such as equipment vibration and sensor interference, and distinguishes transient interference from real persistent bundle breakage diseases. The application realizes hidden bundle breakage bottom-up perception through the change of the mechanical slope, can realize non-blind area identification of hidden area bundle breakage, solves the problems of high false and missed judgment rate, existence of monitoring blind area and insufficient intelligence of the traditional monitoring, realizes rapid early warning and hierarchical treatment of bundle breakage, and effectively improves the safety and monitoring reliability of the prestress tensioning construction.

Description

Technical Field

[0001] This invention belongs to the field of prestressed tensioning technology, specifically relating to a prestressed tensioning tendon breakage identification and early warning system and method. Background Technology

[0002] Prestressed steel strand tensioning is a critical process in bridge, building, tunnel, and municipal geotechnical engineering. If strand breakage or bundle breakage occurs during tensioning, it will lead to insufficient effective prestress, uneven stress on components, and can easily cause major safety accidents such as structural cracking, excessive deformation, or even overall instability and collapse. Therefore, real-time and accurate monitoring of strand breakage during tensioning is a core element in ensuring the construction quality and structural safety of prestressed engineering projects.

[0003] Existing methods for monitoring prestressed tension tendon breakage mainly include three categories: manual visual inspection, fixed-point instrument sampling monitoring, and single-parameter threshold comparison monitoring. Among them, manual monitoring relies heavily on the experience of operators and is affected by on-site dust, light, noise, and obstruction conditions. It is difficult to effectively identify hidden tendon breaks, minor tendon breaks, and instantaneous tendon breakage dynamics, resulting in a high rate of missed detections and significant time lag. Traditional instrument monitoring often uses single strain or displacement parameter exceedance judgments, which cannot characterize the mechanical abrupt change mechanism when tendon breakage occurs. Conventional threshold comparison algorithms only rely on absolute value exceedance warnings and cannot distinguish between actual tendon breakage and data fluctuations caused by construction disturbances, leading to frequent misjudgments and poor reliability.

[0004] In addition, existing technologies lack a collaborative closed-loop mechanism for mechanical judgment and visual verification, resulting in the following inherent defects: First, they cannot effectively detect broken wires in concealed areas, making it difficult to capture broken wire defects in invisible locations such as inside anchors and corrugated pipes, creating monitoring blind spots; second, they cannot effectively distinguish false abnormal data caused by construction disturbances such as sensor electromagnetic interference, equipment mechanical vibration, jack movement, and micro-deformation of supports, leading to low reliability of early warnings; and third, the automation level of the entire process of monitoring, judgment, early warning, and verification is low, data traceability is poor, and it is difficult to meet the requirements of refined and standardized construction supervision. Summary of the Invention

[0005] To address the technical problems existing in the prior art, the first aspect of the present invention provides a prestressed tensioning tendon breakage identification and early warning system. Secondly, based on the same inventive concept, the present invention also provides a prestressed tensioning tendon breakage identification and early warning method based on the aforementioned identification and early warning system, thereby achieving real-time, accurate identification and timely early warning of prestressed tensioning tendon breakage, and improving construction safety and intelligence.

[0006] In this embodiment of the invention, the prestressed strand breakage identification and early warning system includes an information acquisition module, a data transmission module, a storage and calculation module, an early warning handling module, and a verification module. The information acquisition module collects real-time tension force and elongation data throughout the entire tensioning process of a single strand. The data transmission module is electrically connected to the information acquisition module and transmits the force and elongation data collected by the information acquisition module to the storage and calculation module. The storage and calculation module stores the raw data of the entire tensioning process, calculates the dynamic slope of the force-elongation ratio during tensioning, and determines whether there is a suspected strand breakage anomaly based on the comparison between the real-time slope and a preset slope anomaly threshold. The early warning handling module is electrically connected to the storage and calculation module and triggers a suspected early warning when a real-time slope anomaly is detected, and sends a video verification command to the verification module. After receiving the verification result, it performs a confirmation of the early warning or deactivates the suspected early warning. The verification module is electrically connected to the early warning handling module and responds to the video verification command by acquiring a real-time video stream of the tensioning area, identifying the authenticity of the anomaly through the consistency characteristics of the dynamic deformation of the strand tensioning, and outputting a verification result confirming a strand breakage or a misjudgment due to interference.

[0007] The identification and early warning method of this invention, based on the above-mentioned prestressed tensioning tendon breakage identification and early warning system, includes the following steps:

[0008] S1. In the storage and calculation module, the normal slope range of force value-elongation during the tensioning process of each bundle of steel strands is preset, and the slope abnormal threshold is preset based on the normal slope range.

[0009] S2. During the entire tensioning process of the steel strands, the real-time tension force and elongation data of each bundle of steel strands are collected through the information acquisition module;

[0010] S3. The collected tension force and elongation data are uploaded to the storage and calculation module in real time;

[0011] S4. The storage and calculation module calculates the dynamic slope of the tension force value minus the elongation. If the real-time slope exceeds the preset slope abnormality threshold, it is determined that there is a suspected abnormality of bundle breakage, and the early warning and handling module triggers a first-level early warning.

[0012] S5. Start the verification module to collect real-time video stream. The verification module determines whether the actual strand breakage or misjudgment is caused by construction disturbance based on the dynamic characteristics of the steel strand tension.

[0013] S6. After the verification module confirms the breakage, the early warning and handling module triggers a level two early warning, prompting the cessation of tensioning construction; if the verification module determines that it is a misjudgment, the early warning and handling module automatically cancels the level one early warning and prompts the commencement of tensioning construction.

[0014] Compared with the prior art, the advantages of the preferred technical solution of the present invention include:

[0015] 1. This invention employs a dual recognition mechanism combining mechanical slope as the primary criterion and dynamic video verification, abandoning the traditional crude monitoring method that relies on manual experience and visual judgment or simple comparison of single parameters. On one hand, it accurately captures the core mechanical characteristic of the sudden change in stiffness at the moment of strand breakage through the dynamic slope of force-elongation, eliminating missed detections. On the other hand, relying on the consistent characteristics of dynamic deformation of the steel strand, combined with the dual judgment logic of instantaneous slope jump and stable video tensioning shape, it quickly identifies false anomalies caused by construction disturbances such as equipment vibration, sensor interference, jack movement, and micro-deformation of the support, effectively reducing invalid false alarms and completely solving the problems of high misjudgment and missed detection rates in traditional monitoring, significantly improving the accuracy of strand breakage identification and its adaptability to on-site engineering.

[0016] 2. This invention achieves concealed beam break detection by utilizing changes in mechanical slope. It can accurately identify invisible beam breaks in areas obscured by anchors, corrugated pipes, etc., without the need for direct observation of the break point, thus eliminating monitoring blind spots. At the same time, it integrates five types of quantitative video features, including inter-frame displacement, beam deformation, vibration energy, single-filament springback, and strain unevenness, to conduct multi-dimensional discrimination. Compared with a single discrimination method, the identification standard is more refined and scientific, which can accurately distinguish between real beam breaks and construction interference anomalies, improving the stability and accuracy of beam break identification in complex construction sites.

[0017] 3. This invention constructs a multi-feature quantitative fusion intelligent discrimination model, abandoning the traditional qualitative and experience-based manual discrimination mode. It integrates five core bundle breakage features through normalized weighted fusion of force value drop, bundle jitter energy, local displacement mutation, single filament springback relaxation, and strain unevenness. Combined with the 3σ statistical criterion and sliding window dual verification mechanism, it achieves standardized and quantitative judgment of bundle breakage anomalies. At the same time, it can adaptively optimize weights and thresholds according to different specifications of steel strands and different construction conditions, and has strong adaptability to working conditions and algorithm robustness.

[0018] 4. This invention addresses various instantaneous construction disturbances at construction sites by quickly and accurately distinguishing between construction disturbances and actual beam breakages through interference-specific features such as instantaneous slope jumps, normal video morphology, and recoverable force values, and automatically deactivating invalid warnings. At the same time, the system can automatically record various interference data, iteratively optimize the discrimination model, and continuously reduce the frequency of similar invalid warnings. This ensures that beam breakage hazards are not missed and construction safety is controllable, while avoiding frequent unnecessary warnings that interfere with normal tensioning construction, effectively improving project adaptability and construction efficiency.

[0019] 5. This invention achieves intelligent and digital closed-loop management of the entire prestressed tensioning tendon breakage monitoring process. It integrates high-frequency synchronous data acquisition, encrypted secure transmission, intelligent algorithm quantitative discrimination, hierarchical precise early warning, video visualization verification, data traceability and archiving, and model adaptive optimization functions to form a complete intelligent monitoring system. This system can quickly respond to tendon breakage anomalies, enabling early detection, early warning, and early handling of potential hazards, effectively avoiding major safety risks such as cracking, instability, and collapse of prestressed structures. Simultaneously, it retains traceable monitoring data throughout the entire process, providing reliable data support for project quality acceptance, construction process iteration and optimization, and safety responsibility traceability, comprehensively improving the standardization and intelligent management level of prestressed engineering construction. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the prestressed tensioning tendon breakage identification and early warning system of an embodiment. Detailed Implementation

[0021] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0022] Example 1

[0023] This embodiment provides a prestressed tensioning tendon breakage identification and early warning system (hereinafter referred to as the tendon breakage identification and early warning system), such as Figure 1 As shown, in a preferred embodiment, it includes an information acquisition module, a data transmission module, a storage and computing module, an early warning and handling module, and a verification module.

[0024] The information acquisition module is used to collect real-time tension force and elongation data throughout the entire tensioning process of a single strand of steel wire. The module includes a force sensor and an elongation sensor. Each strand of prestressed steel wire is equipped with an independent set of sensors to eliminate data interference between strands. The force sensor collects the tension force of each strand, and the elongation sensor collects the elongation of each strand. The signal outputs of the force sensor and the elongation sensor are connected to the input of the storage and calculation module. The force sensor is a hollow shaft pressure sensor, coaxially mounted between the tensioning jack and the anchor plate. The bearing surface is perpendicular to the tensioning axis of the steel strand, fully bearing the tensioning load of the jack. The force sensor is a high-precision tension sensor with a measurement accuracy of ±0.5%FS. The elongation sensor is fixed on an independent reference support of the beam. The laser emitter is aligned with the end face of the jack piston rod to avoid obstruction by the movement of the equipment. The elongation of the steel strand is equivalently converted by the extension of the piston rod. The elongation sensor is a laser displacement sensor with a measurement accuracy of ±0.1mm. The data acquisition frequency is 10Hz~50Hz, realizing continuous sampling of dynamic data throughout the tensioning process.

[0025] The data transmission module is electrically connected to the information acquisition module and is used to transmit the force and elongation data collected by the information acquisition module to the storage and computing module. The data transmission module adopts a 5G, WiFi, or LoRa wireless transmission architecture and is equipped with an AES data encryption unit to encrypt the real-time force and elongation data during transmission, preventing data loss or tampering caused by electromagnetic interference at the construction site and ensuring the real-time performance and security of data transmission.

[0026] The storage and calculation module stores the raw data of the entire tensioning process, calculates the dynamic slope of force-elongation during tensioning, and determines whether there are any suspected strand breakage anomalies based on the comparison between the real-time slope and the preset slope anomaly threshold. The storage and calculation module includes a data storage unit and a slope calculation unit. The data storage unit adopts a dual backup mode of local solid-state drive + cloud server to achieve temporary storage of real-time data and long-term traceability archiving. The slope calculation unit uses a real-time iterative algorithm to calculate the ratio of the elongation difference to the force difference at adjacent sampling times frame by frame to obtain the dynamic tensioning slope. The storage and calculation module can also adaptively calibrate the normal slope range according to the steel strand specifications and elastic modulus parameters, dynamically update the anomaly threshold, and adapt to different construction conditions.

[0027] The early warning and handling module is electrically connected to the storage and computing module. It triggers a suspected early warning when an anomaly in the real-time slope is detected, and sends a video verification command to the verification module. After receiving the verification result, it either confirms the early warning or cancels the suspected early warning. The early warning and handling module has a two-level early warning mechanism: a primary warning and a secondary warning. The primary warning is a suspected beam breakage warning, which only triggers a mild on-site audio-visual alert to remind workers to pay attention to the working condition. The secondary warning is a confirmed beam breakage warning after confirmation by the verification module, which triggers a high-decibel audio-visual alarm and pushes a shutdown command to the worker's mobile device to eliminate safety risks.

[0028] The verification module is electrically connected to the early warning and handling module. It responds to video verification commands, collects real-time video streams of the tensioning area, and identifies the authenticity of anomalies based on the consistency characteristics of the dynamic deformation of the steel strands during tensioning. It outputs verification results confirming broken strands or misjudgments due to interference. The verification module includes multiple high-definition cameras and a video analysis unit. The high-definition cameras fully cover the tensioning area of ​​each steel strand bundle. For example, each monitored steel strand bundle is equipped with two 4K@30fps cameras. The main camera is obliquely positioned above the tensioning end, fully covering the anchorage, exposed steel strands, and jacks. The auxiliary camera is positioned on the side of the beam to collect deformation data of the web and bottom plate, capturing the tensioning video stream without blind spots, including sequential images of the steel strand tensioning process. The images collected by the main and auxiliary cameras are calibrated using a checkerboard calibration method for internal and external parameters. The two video streams are time-series aligned based on timestamps, and the feature points extracted from both sets of images are uniformly mapped to the same world coordinate system before joint calculations are performed.

[0029] The video analysis unit extracts inter-frame displacement, bundle deformation, and vibration characteristics during the tensioning process of the steel strands using an inter-frame difference algorithm. Without needing to identify specific fracture points, it can determine the authenticity of anomalies based on the overall consistency of the tensioning state, and is also suitable for monitoring and verifying concealed strand breaks. Specifically, the video analysis unit calculates the inter-frame displacement field of the steel strand bundle frame by frame, obtaining the real-time displacement and displacement rate at each point during tensioning; it extracts the overall deformation uniformity characteristics of the steel strand bundle, identifying anomalous deformations such as sudden local tensile changes, bundle bending, and relaxation bulging; and it extracts the vibration amplitude, frequency, and duration of vibration during tensioning to distinguish between instantaneous abnormal vibrations and mechanical imbalance vibrations. The video analysis unit employs the ORB feature point tracking algorithm. Pre-defined artificial reference feature points are established for the steel strands, anchor plates, and beam structure. After extracting feature points from the anchorage and steel strand surfaces using the ORB algorithm, a motion threshold is set to filter two types of invalid feature points: first, feature points whose movement speed exceeds the steel strand tensioning rate range, identified as interference features from construction personnel or moving equipment; second, feature points whose spatial location exceeds the pre-defined ROI (Region of Interest) for the tensioning operation, which are directly discarded. The video analysis unit also incorporates a static background noise reduction algorithm. The first N frames of the video are used to construct a static reference background. Subsequent frames are then compared with this reference background to eliminate pixel interference from static backgrounds such as the jack frame and the fixed beam structure, retaining only the dynamic features generated by the steel strand deformation.

[0030] In this invention, the broken strand identification and early warning system also includes a display module. The display module adopts an industrial-grade touch screen to display the tension value, elongation, dynamic slope, early warning status and verification results of each strand of steel strand in real time. It supports historical data curve playback and traceability query, which makes it convenient for on-site personnel to intuitively monitor the construction status.

[0031] Example 2

[0032] This embodiment provides a method for identifying and warning of prestressed tendon breakage, based on the prestressed tendon breakage identification and warning system of Embodiment 1, and includes the following steps:

[0033] S1. In the storage and calculation module, a normal slope range for force-elongation during the tensioning process of each steel strand is preset, and an abnormal slope threshold is preset based on this normal slope range. The abnormal slope threshold is adaptively adjusted according to the specifications and elastic modulus parameters of the steel strand. The normal slope range is calibrated through preliminary tests, and the abnormal slope threshold is determined by combining tensioning data from similar projects. For example, if a commonly used steel strand with a diameter of 15.2mm and an elastic modulus of 195GPa is selected for parameter calibration, and the force-elongation slope range under normal tensioning is determined to be 0.005~0.008mm / kN through preliminary tensioning tests, the upper limit of the range is increased by 18% to set the abnormal threshold at 0.0094mm / kN. At the same time, a video acquisition resolution of 4K and a frame rate of 30fps are preset, and graded early warning and handling rules are configured.

[0034] For each specification of steel strand, select no fewer than 20 sets of specimens to carry out standard elastic tension tests, calculate the distribution range of force-elongation slope, and thus define the normal slope range of normal tension for that type of steel strand.

[0035] S2. During the entire tensioning process of the steel strands, the information acquisition module continuously and synchronously acquires the real-time tension force and elongation data of each bundle of steel strands at a frequency of 30Hz to ensure that the force and displacement data are aligned in time.

[0036] S3. The collected tension and elongation data are uploaded to the storage and computing module in real time. The local unit caches the data in real time, and the cloud synchronously backs up and archives the data to ensure that the data is traceable throughout the process.

[0037] S4. The storage and calculation module calculates the dynamic slope of the tension value minus the elongation. If the real-time slope exceeds the preset slope abnormality threshold, that is, when the real-time slope exceeds the threshold of 0.0094mm / kN, it is determined that there is a suspected abnormality of bundle breakage, and the early warning and handling module triggers a first-level early warning.

[0038] S5. The system automatically starts the verification module to collect real-time video streams. The verification module determines whether the breakage is real or a misjudgment caused by construction disturbance based on the dynamic characteristics of the steel strand tensioning. Specifically, if the overall stiffness of the steel strand permanently decreases, the slope continuously exceeds the limit, and there are mechanical imbalance characteristics such as strand jitter, sudden changes in local displacement, single filament rebound relaxation, and uneven deformation, it is determined to be a real breakage, including hidden breaks within the anchor or pipe. If only a single frame shows an instantaneous slope jump, and the steel strand tensioning in the video is smooth throughout, with uniform deformation and no dynamic changes between frames, it is determined to be a misjudgment caused by construction disturbances such as equipment vibration, sensor interference, jack movement, and micro-deformation of the support.

[0039] S6. After the verification module confirms the breakage, the early warning and handling module triggers a level-two early warning, prompting the cessation of tensioning construction. If the verification module determines it is a false alarm, the early warning and handling module automatically cancels the level-one early warning (i.e., cancels the invalid early warning), prompts the commencement of tensioning construction, records the current interference condition, and iteratively optimizes the threshold model to reduce the probability of similar false alarms in the future. All monitoring, early warning, verification, and handling data are archived and saved throughout the process.

[0040] This invention adopts a two-level serial progressive discrimination architecture, rather than two independent schemes: Level 1: Force value-elongation dynamic slope is used for preliminary coarse screening to quickly lock suspected broken bundles and reduce video computing power consumption; Level 2: Only when the Level 1 warning is triggered, the multi-feature fusion model is activated to perform the final verification of authenticity.

[0041] In another preferred embodiment of the present invention, the verification module is connected to the storage and computing module through the data transmission module. The storage and computing module receives multi-feature data uploaded by the verification module. The storage and computing module is configured with a multi-feature fusion-based beam breakage identification algorithm to determine suspected beam breakage anomalies, including the following steps:

[0042] S51. Multiple features must include at least the force drop feature. Beam jitter energy characteristics Local displacement abrupt change characteristics Monofilament resilience and relaxation characteristics and strain non-uniformity characteristics After normalizing the above features, a weighted fusion was performed, and each feature was normalized to the [0,1] interval to construct a comprehensive evaluation index for bundle breakage. ;

[0043] in, , , , , The weight coefficients are all ≥ 0 and satisfy the following conditions: In this embodiment, the optimized value is: , , , , .

[0044] In this embodiment, the initial values ​​of the weight coefficients are obtained through the Analytic Hierarchy Process (AHP). Specifically, the following steps are taken: a pairwise comparison judgment matrix is ​​constructed for five types of wire breakage characteristics: force drop, beam vibration kinetic energy, sudden change in local displacement, single wire springback relaxation, and uneven strain. Several senior engineers in the field of bridge prestressing are invited to score the importance of each characteristic in representing the failure of steel strand wire breakage. The consistency of the scored judgment matrix is ​​checked, and invalid scoring samples with inconsistencies exceeding the standard are removed. The matrix eigenvectors are solved to obtain the initial weights of each characteristic. Then, the initial weights are iteratively corrected based on a large number of wire breakage samples and normal tensioning samples from the construction site, and finally, the optimized weight values ​​are obtained.

[0045] S52, Comprehensive Evaluation Index for When Bundles Are Broken Instantaneous exceedance of threshold At that time, it was determined to be a suspected beam break, triggering a Level 1 warning (instantaneous trigger):

[0046]

[0047] in, Determined by combining statistical results of normal tension data with the 3σ criterion: 3σ, Under normal operating conditions The mean, σ is the standard deviation. Threshold The calibration method is as follows: 30 sets of standard steel strand specimens of different specifications and under different tensioning conditions were selected to carry out normal tensioning tests without wire breakage, and the comprehensive evaluation index of wire breakage under the entire time sequence was collected. ; Calculate the mean μ and standard deviation σ of all sample data according to statistical methods. 3σ determines the first-level early warning threshold to avoid false triggering caused by random fluctuations in normal operating conditions.

[0048] S53. This invention uses a unified convention for global time-series parameters, which is shared by all algorithms in the entire paper. Within the time window [ , ]Inside, At the current sampling time, The sliding time window length is used to confirm a true beam breakage and trigger a level 2 warning (continuous abnormality + irreversible fall) if both of the following conditions are met simultaneously:

[0049] 1) Time window [ , The cumulative amount of the comprehensive evaluation index for internal bundle breakage meets the standard: > , for[ , Small units within ]; among which, The threshold value for the cumulative index between normal operating conditions and beam breakage conditions is determined by calibrating the boundary between beam breakage failure samples and normal operating condition samples. The calibration method is as follows: A boundary test is carried out with a control group. One group consists of failed steel strand samples with artificially pre-fabricated broken wires, and the other group consists of defect-free normally tensioned samples. The time-series integral results of the comprehensive index of the two types of samples are collected, and the median boundary value of the integral values ​​of the two types of samples is taken as the boundary threshold. This is used to distinguish between real wire breakage and instantaneous disturbance.

[0050] 2) Tension does not recover, i.e., the breakage is irreversible, and the force value falls irreversibly due to the constraint: ;in, The prestress relaxation tolerance threshold is determined by long-term prestress relaxation tests on steel strands of the same specification. Only when the tension drop exceeds the normal relaxation range is it considered an irreversible strand breakage failure. Based on multiple sets of long-term prestressed relaxation tests on steel strands of the same specification, the irreversible force drop limit of the steel strands under long-term load is collected. Combined with the prestress loss limit allowed by bridge specifications, the irreversible tension drop threshold is determined. .

[0051] For the length of the sliding time window The system adaptively adjusts the length of the sliding time window based on three core parameters: the nominal cross-sectional area of ​​the steel strand, the elastic modulus, and the preset tensioning rate. The larger the cross-sectional area and the slower the tensioning rate, the longer the window duration will be to ensure that the changing trends of force and deformation can be fully captured; the smaller the cross-sectional area and the faster the tensioning rate, the shorter the window duration will be to improve the response speed of anomaly identification.

[0052] In this invention, the force drop feature To balance instantaneous collapse and gradual slippage, a sliding window cumulative calculation method is used:

[0053]

[0054]

[0055] in, This is the current sampling time; The sampling time step is equal to the sampling period; The sliding time window length is typically 0.2–1.0 s in engineering practice and can be adaptively adjusted according to the steel strand specifications and tensioning rate. This represents the change in tension force within one sampling step. The tension force at the current sampling moment, i.e., the real-time tension force, is acquired by the force sensor; The tension at the previous sampling time; for[ , Small units within ]

[0056] In this invention, the beam jitter energy characteristics The beam vibration state is calculated based on the displacement of characteristic points and normalized according to the tension amplitude:

[0057]

[0058]

[0059]

[0060]

[0061] Where N is the number of feature points on the surface of the steel strand; i is the i-th feature point on the surface of the steel strand, 1≤i≤N; for[ , Small units within ]; subscripts of the above parameters This represents the overall average parameters of the steel strand;

[0062] The displacement of the i-th feature point is obtained by tracking the video sequence images. The specific calculation method is as follows:

[0063] The pixel coordinates of the i-th feature point in the initial frame are ( );

[0064] The pixel coordinates of the i-th feature point at time t are ( );

[0065] Pixel offset in tension direction: ;

[0066] Converted to physical displacement: ;

[0067] in, The pixel-to-physical length conversion factor obtained from camera calibration, in mm / pixel, is the pixel coordinate difference × This can be converted to actual physical displacement in mm, using a pixel-to-physical length conversion factor. The standard checkerboard calibration method is used to solve the problem. The calibration plate is placed on the tensioning operation plane to complete the camera calibration and obtain the actual millimeter length corresponding to each pixel.

[0068] It should be noted that the average displacement of video feature points is equivalent to the overall elongation of a single bundle of steel strands, which can also be obtained by an elongation sensor.

[0069] In this invention, the feature of sudden local displacement is... The displacement gradient change rate is calculated based on the displacement difference between adjacent feature points and then globally normalized.

[0070]

[0071]

[0072]

[0073] in, This represents the displacement of the (i-1)th feature point. When i=1, the fixed end is taken as the reference. The reference point is selected at the fixed geometric center of the anchor plate at the anchoring end of the steel strand, and is used as the origin of the coordinates for displacement calculation. The initial physical distance between adjacent feature points; yes The derivative with respect to time.

[0074] In this invention, the monofilament elastic relaxation feature To identify the reverse springback behavior of broken wires and correct the sensitivity decay problem in the later stage of tensioning, the following method was used:

[0075]

[0076]

[0077] in, This refers to the springback displacement. The displacement of the i-th feature point; The displacement of the i-th feature point at the previous sampling time; For a very small constant, for example, take This is used to prevent calculation errors caused by a denominator of 0. The value is much smaller than the actual measured displacement in the engineering and will not affect the characteristic calculation results.

[0078] In this invention, the strain non-uniformity feature The strain dispersion caused by stress redistribution due to bundle breakage is obtained in the following way:

[0079]

[0080]

[0081] in, The strain of the i-th local region is calculated using the displacement difference between adjacent feature points: When i=1, the fixed end is taken as the reference. .

[0082] For edge devices on construction sites with limited computing resources, the multi-feature model can be downgraded and simplified: remove two types of features with high computing overhead, such as beam vibration kinetic energy and uneven strain, and retain only three core features, such as force value drop, sudden change in local displacement, and single-filament springback relaxation, to complete the fusion judgment; or it can be directly downgraded to a simple solution of pre-slope single index judgment + video verification, which is suitable for low computing hardware scenarios.

[0083] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A prestressed tensioning tendon breakage identification and early warning system, characterized in that, It includes an information collection module, a data transmission module, a storage and computing module, an early warning and response module, and a verification module; The information acquisition module is used to collect real-time tension force and real-time elongation data of a single strand of steel strand throughout the entire tensioning process; The data transmission module is electrically connected to the information acquisition module and is used to transmit the force data and elongation data acquired by the information acquisition module to the storage and calculation module. The storage and calculation module is used to store the raw data of the entire tensioning process, calculate the dynamic slope of force value-elongation during the tensioning process, and determine whether there is a suspected abnormality of bundle breakage based on the comparison result of the real-time slope and the preset slope abnormality threshold. The early warning processing module is electrically connected to the storage and computing module. It is used to trigger a suspected early warning when the real-time slope is determined to be abnormal, and to send a video review instruction to the review module. After receiving the review result, it performs the operation of confirming the early warning or canceling the suspected early warning. The verification module is electrically connected to the early warning and handling module. It is used to respond to video verification commands, collect real-time video streams of the tensioning area, identify the authenticity of abnormalities through the consistency characteristics of the dynamic deformation of the steel strand tensioning, and output verification results confirming the breakage or interference misjudgment.

2. The prestressed tensioning tendon breakage identification and early warning system according to claim 1, characterized in that, The information acquisition module includes a force sensor and an elongation sensor corresponding to each bundle of steel strands. The force sensor is used to acquire the tension force value of each bundle of steel strands, and the elongation sensor is used to acquire the elongation of each bundle of steel strands. The signal output terminals of the force sensor and the elongation sensor are connected to the input terminal of the storage and calculation module. And / or the verification module includes a high-definition camera and a video analysis unit. The high-definition camera covers the tensioning operation area of ​​each strand of steel strand, and the video analysis unit is used to extract inter-frame displacement, strand deformation and vibration characteristics during the tensioning process to realize dynamic anomaly identification under non-visual fracture conditions of the steel strand.

3. A method for identifying and warning of prestressed tension tendon breakage, characterized in that, Based on the prestressed tensioning breakage identification and early warning system described in claim 1 or 2, the system includes the following steps: S1. In the storage and calculation module, the normal slope range of force value-elongation during the tensioning process of each bundle of steel strands is preset, and the slope abnormal threshold is preset based on the normal slope range. S2. During the entire tensioning process of the steel strands, the real-time tension force and elongation data of each bundle of steel strands are collected through the information acquisition module; S3. The collected tension force and elongation data are uploaded to the storage and calculation module in real time; S4. The storage and calculation module calculates the dynamic slope of the tension force value minus the elongation. If the real-time slope exceeds the preset slope abnormality threshold, it is determined that there is a suspected abnormality of bundle breakage, and the early warning and handling module triggers a first-level early warning. S5. Start the verification module to collect real-time video stream. The verification module determines whether the actual strand breakage or misjudgment is caused by construction disturbance based on the dynamic characteristics of the steel strand tension. S6. After the verification module confirms the breakage, the early warning and handling module triggers a level two early warning, prompting the cessation of tensioning construction; if the verification module determines that it is a misjudgment, the early warning and handling module automatically cancels the level one early warning and prompts the commencement of tensioning construction.

4. The identification and early warning method according to claim 3, characterized in that, In step S1, the slope anomaly threshold is adaptively adjusted according to the specifications and elastic modulus parameters of the steel strand, the normal slope range is calibrated through previous tests, and the slope anomaly threshold is determined in combination with tensioning data from similar projects. In step S5, the verification module determines whether the actual strand breakage or misjudgment is caused by construction disturbance based on the dynamic characteristics of the steel strand tensioning as follows: If the overall stiffness of the steel strand permanently decreases, the slope continuously exceeds the limit, and there are mechanical imbalance characteristics such as strand jitter, sudden changes in local displacement, single filament springback relaxation, and uneven deformation, then it is determined to be an actual strand breakage; if only a single frame instantaneous slope jump occurs, and the steel strand tensioning in the video is smooth throughout, with uniform deformation and no dynamic changes between frames, then it is determined to be a misjudgment caused by construction disturbance.

5. The identification and early warning method according to claim 3 or 4, characterized in that, The storage and computing module is equipped with a multi-feature fusion-based beam breakage detection algorithm to determine suspected beam breakage anomalies, including the following steps: S51, the multiple features include at least the force drop feature. Beam jitter energy characteristics Local displacement abrupt change characteristics Monofilament resilience and relaxation characteristics and strain non-uniformity characteristics After normalizing the above features, a weighted fusion was performed to construct a comprehensive evaluation index for bundle breakage. ; in, , , , , The weight coefficients are all ≥ 0 and satisfy the following conditions: ; S52, Comprehensive Evaluation Index for When Bundles Are Broken Instantaneous exceedance of threshold At that time, it was determined to be a suspected beam breakage, triggering an alert: in, Determined by combining statistical results of normal tension data with the 3σ criterion: 3σ, Under normal operating conditions The mean of σ is the standard deviation; S53, within the time window [ , ] Inside, At the current sampling time, The sliding time window length is used to confirm a true fasciculation if both of the following conditions are met: The cumulative amount of comprehensive evaluation indicators for bundle breakage within the time window meets the standard: > , for[ , Small units within ]; among which, The cumulative index threshold between normal operating conditions and beam breakage conditions is obtained by calibration from experimental data. The tension does not recover, indicating the irreversible nature of the broken bundle: ;in, The prestress relaxation tolerance threshold is determined by long-term prestress relaxation tests on steel strands of the same specification.

6. The identification and early warning method according to claim 5, characterized in that, The force drop characteristic To balance instantaneous collapse and gradual slippage, a sliding window cumulative calculation method is used: in, This is the current sampling time; This is the sampling time step; The length of the sliding time window; This represents the change in tension force within one sampling step. The tension at the current sampling moment; The tension at the previous sampling time; for[ , Small units within ] 7. The identification and early warning method according to claim 6, characterized in that, The beam jitter energy characteristics The beam vibration state is calculated based on the displacement of characteristic points and normalized according to the tension amplitude: Where N is the number of feature points on the surface of the steel strand; i is the i-th feature point on the surface of the steel strand, 1≤i≤N; the subscripts of the above parameters This represents the overall average parameters of the steel strand; for[ , Small units within ]; The displacement of the i-th feature point is obtained by tracking the video sequence images. The specific calculation method is as follows: The pixel coordinates of the i-th feature point in the initial frame are ( ); The pixel coordinates of the i-th feature point at time t are ( ); Pixel offset in tension direction: ; Converted to physical displacement: ; in, The pixel-to-physical length conversion factor obtained from camera calibration.

8. The identification and early warning method according to claim 7, characterized in that, The local displacement abrupt change feature The displacement gradient change rate is calculated based on the displacement difference between adjacent feature points and then globally normalized. in, This represents the displacement of the (i-1)th feature point. When i=1, the fixed end is taken as the reference. ; The initial physical distance between adjacent feature points; yes The derivative with respect to time.

9. The identification and early warning method according to claim 8, characterized in that, The monofilament's elastic relaxation characteristic To identify the reverse springback behavior of broken wires and correct the sensitivity decay problem in the later stage of tensioning, the following method was used: in, This refers to the springback displacement. The displacement of the i-th feature point; The displacement of the i-th feature point at the previous sampling time; It is a very small constant used to prevent the denominator from being 0.

10. The identification and early warning method according to claim 9, characterized in that, The strain non-uniformity feature The strain dispersion caused by stress redistribution due to bundle breakage is obtained in the following way: in, The strain of the i-th local region is calculated using the displacement difference between adjacent feature points: When i=1, the fixed end is taken as the reference. .

Images