A self-adaptive tensioning control method for a bonded prestressed beam based on a BP neural network

Abstract

The application discloses a kind of based on BP neural network's slow-bonding prestressed beam self-adapting tension control method, belong to building engineering technical field.The method includes: constructing and training BP neural network prediction model, the mapping relationship of upper construction load, beam bottom edge stress and tension control force is established;Obtain structural parameters and load information, set load control factor initial value and increment step;Start iteration optimization program, dynamically update upper construction load value and input model output actual execution tension force;Through load control factor, judge iteration termination condition;Based on the tension control force calculated by output, the beam bottom edge stress is calculated and the anti-cracking safety check is carried out, and the alarm is triggered when it is not satisfied.The application trains BP neural network prediction model, discards experience estimation, realizes high-precision self-adaptive control of tension force;Build intelligent prediction and physical check double security system and whole-process alarm processing mechanism, apply the characteristics that slow-bonding prestressed can be tensioned multiple times before binder solidification, effectively solve the problem that traditional method tension control power process is complex, lack of dynamic response and safety protection is insufficient, effectively control structural crack in prestressed tension engineering, significantly improve the safety of structure construction.

Description

Technical Field

[0001] This invention relates to the field of building engineering technology, and in particular to an adaptive tensioning control method for loosely bonded prestressed beams based on a BP neural network. Background Technology

[0002] In the field of building construction, loosely bonded prestressed beams are widely used in large-span structures due to their combination of the ease of construction of unbonded prestressed beams and the excellent stress performance of bonded prestressed beams. However, in actual construction, crack control of prestressed beams faces severe challenges: Existing prestressing tension control methods mostly rely on empirical formulas or static calculations, making it difficult to accurately quantify the nonlinear impact of dynamic changes in the superstructure load on the beam's stress. Traditional "estimation-based" control methods often result in inaccurate prestress quantification and an inability to respond in real time to complex working conditions during construction.

[0003] As construction progresses, the superstructure load increases dynamically. Traditional methods lack the ability to perceive and adaptively adjust to such dynamic load changes in real time, which can easily lead to a mismatch between tension and actual requirements. When the tension is insufficient, the tensile stress at the bottom edge of the beam exceeds the tensile strength limit of the concrete, directly causing cracks in the structure; when the tension is excessive, it may cause excessive camber or other structural damage.

[0004] In traditional construction processes, crack resistance safety verification often lags behind tensioning operations, lacking a real-time prediction, verification, and alarm closed-loop mechanism. Once the stress at the bottom edge of the beam exceeds the safety threshold, it is often only discovered after cracks appear, at which point remedial costs are high and the structural durability and safety are seriously affected.

[0005] Therefore, a control technology is needed that can establish a high-precision mapping relationship between load, stress and tension, fully utilize the characteristic of slow-bonding prestress that it can be tensioned multiple times before the adhesive cures, and can adaptively adjust the tension control force according to real-time working conditions and have an instant anti-cracking safety early warning function, so as to fundamentally solve the problem of structural crack control in the construction of prestressed beams. Summary of the Invention

[0006] This invention provides an adaptive tensioning control method for loosely bonded prestressed beams based on a BP neural network. This method addresses the technical problems in traditional technologies, such as insufficient accuracy due to reliance on empirical estimation for prestress quantification, inability to respond in real time to dynamic changes in construction loads and structural nonlinear characteristics, and a lack of timely early warning capabilities due to delayed crack resistance safety verification. These issues make it difficult to effectively suppress the generation of structural cracks.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: An adaptive tension control method for loosely bonded prestressed beams based on a backpropagation neural network includes the following steps: S1: Construct and train a BP neural network prediction model to establish the mapping relationship between the upper construction load, the stress at the bottom edge of the beam and the tension control force; S2: Obtain the structural parameters of the prestressed beam to be constructed and the current superstructure construction load information, and set the load control factor. Initial values, initial values ​​of upper construction loads, stress values ​​at the bottom edge of the beam, and increment step size. Among them, the stress value at the bottom edge of the beam is the limit of the concrete design tensile strength and the safety factor. The product; S3: Start the iterative optimization program. In each iteration, increment the step size. Cumulative update of load control factors Determine whether the load control factor is satisfied. If the condition is not met, proceed to step S4; if the condition is met, proceed to step S5. S4: Calculate the new upper construction load value as the product of the initial upper construction load value and the load control factor. Input the updated upper construction load value and the stress value at the bottom edge of the beam into the BP neural network prediction model, output the actual tension force under the current working condition, and return to step S3 to continue the iteration. S5: Order The process ends after inputting the initial value of the upper construction load and the stress at the bottom edge of the beam into the BP neural network prediction model and outputting the final actual tension control force. S6: Based on the actual tension force output in steps S4 and S5, calculate the edge stress at the bottom of the beam under the action of the actual tension force. Determine the edge stress Does it meet the crack resistance safety constraint conditions? ,in This is the standard value for the axial tensile strength of concrete; if it is not met, an alarm will be triggered.

[0008] In step S1, the training process of the BP neural network prediction model includes: S1.1: Construct a dataset containing the superstructure construction load, the bottom edge stress of the beam, and the tension control force; the tension control force is calculated from the elongation value of the prestressing tendon at the tensioning end, wherein the calculation process takes into account the friction loss of the loosely bonded prestressing tendon on the tensioning path to compensate for the tension control force; the bottom edge stress of the beam is calculated by structural mechanics formulas or obtained by field measurement; S1.2: Normalize the data and construct a BP neural network with at least one hidden layer, wherein the hidden layer contains several neurons; the activation function between the input layer and the hidden layer adopts the Poslin positive linear transfer function, and the activation function between the hidden layer and the output layer adopts the Purelin linear transfer function. S1.3: Divide the dataset into a training set, a validation set, and a test set, set a fixed ratio and learning rate, and train the dataset using the upper construction load and the stress at the bottom edge of the beam as input values ​​and the tension control force as output values. S1.4: Output the mean squared error curves of the training set, validation set, and test set during the training process of the BP neural network model, until the mean squared error is less than the preset accuracy threshold, and the model is judged to have been successfully trained.

[0009] In step S2, the load control factor The method for determining the initial value is as follows: If it is the first stage of construction under the first level of load. The initial value is set to 0; If it is the subsequent... During the construction phase under high load, The formula for calculating the initial value is: ,in, This represents the cumulative load of the previous level. This is the cumulative design load for the current level.

[0010] In step S2, the incremental step size Δη is an adaptive adjustment value, and its calculation formula is as follows: Formula ①, Formula ②, Formula ③, Formula ④, and Formula ⑤: ① ② ③ ④ ⑤ In the formula, ; For span and section height coefficients, For the beam span, The height of the beam section. For concrete performance coefficients, This refers to the standard value of the compressive strength of a concrete cube. The elastic modulus of concrete. This is the prestressed tendon reinforcement ratio coefficient. This refers to the prestressed tendon reinforcement ratio. For stress level and safety reserve factor, This represents the allowable tensile stress control value for the beam. This is the standard value for the axial tensile strength of concrete.

[0011] Pressure sensors are installed in the upper construction area of ​​the beam to collect upper construction load data. At least three sensors are installed in each span of the beam and are evenly distributed in the span direction of the beam. Strain gauges are attached to the bottom center of the beam and the bottom of the beam side mid-span to collect beam strain data. At least three strain gauges are installed at each location.

[0012] In step S6, after the edge stress is detected to be not in compliance with the crack resistance safety constraint, an alarm action is automatically triggered within 1 second, and processing measures and alarm recording and tracing actions are performed. The alarm details are entered into the database and a ledger is formed.

[0013] The alarm action includes issuing an audible and visual warning signal through an audible and visual alarm device and displaying it visually on the control terminal. The sound and light warning signals are emitted from locations covering the tensioning control terminal and the tensioning operation area at the construction site. The warning methods include flashing high-brightness LED warning lights and high-decibel buzzer sounds. The visualization is executed on the control terminal integrated with the tension control system. After the alarm is triggered, a red alarm pop-up window will automatically pop up, displaying the alarm type, alarm time, and current measured stress value.

[0014] The processing measures include: Emergency intervention measures: Suspend tensioning operations, evacuate on-site personnel, and collect real-time monitoring data at high frequency; Inspection and rectification measures: Conduct inspections on excessive upper construction loads, prestressed tendon tension status, sensor data accuracy, and beam condition, and then implement rectification measures. Resumption of construction measures: Perform hazard verification, alarm reset, and issue a resumption of construction order.

[0015] The present invention provides an adaptive tension control method for loosely bonded prestressed beams based on a BP neural network, which has the following significant advantages: This invention, by constructing and training a BP neural network prediction model, abandons traditional empirical estimation and establishes a high-precision nonlinear mapping relationship between the superstructure construction load, the stress at the bottom edge of the beam, and the tension control force. Based on real-time collected superstructure loads and set crack-resistant stress limits, iteratively optimizes the program to dynamically output the optimal actual tension force. This adaptive control ensures that the compressive stress provided by the prestressing tendons accurately counteracts the tensile stress generated by the external load, strictly controlling the stress at the bottom edge of the beam within the safe range of the concrete's tensile strength, effectively preventing structural cracks caused by improper tension.

[0016] This invention introduces an adaptive incremental step size algorithm that comprehensively considers multiple factors such as beam span, cross-sectional height, concrete properties, reinforcement ratio, and stress level. This enables the method to adapt to complex working conditions of beams with different specifications and at different construction stages, ensuring that it can provide the optimal crack-resistant tensioning scheme under various dynamic changes, thus solving the problem of traditional methods lacking dynamic response capabilities.

[0017] This invention constructs a dual safety assurance system of intelligent prediction and physical verification. It not only utilizes a BP neural network for feedforward prediction, directly outputting the actual tension force that takes into account the friction loss of the reinforcing steel, but also, after outputting the tension force, calculates the stress at the bottom edge of the beam based on mechanical formulas and compares it in real time with the standard value of the axial tensile strength of concrete. Tensioning is only executed when the calculation result meets the crack resistance requirements; otherwise, an alarm is immediately triggered. This dual verification mechanism greatly improves the reliability of crack resistance control, eliminates the error risks that may exist in single-model prediction, and ensures that the structure is always in a crack-free and safe working state.

[0018] This invention features a real-time monitoring and immediate alarm mechanism throughout the entire process. Once it detects that the edge stress does not meet the crack resistance safety constraints, it can automatically trigger an audible and visual alarm within one second and execute emergency intervention measures such as suspending tensioning, evacuating personnel, and initiating high-frequency monitoring. Simultaneously, it provides detailed troubleshooting and rectification guidelines, forming a closed-loop management system encompassing monitoring, alarming, handling, and resetting, significantly improving the structural safety during construction.

[0019] This invention solves the technical problems in traditional technologies where prestress quantification relies on empirical estimation, resulting in insufficient accuracy; inability to respond in real time to dynamic changes in construction loads and structural nonlinear characteristics; and a lack of immediate early warning capability due to delayed crack resistance safety verification, thus making it difficult to effectively suppress the generation of structural cracks. Other features and advantages of this invention will be set forth in the following description and will be apparent in part from the description, or may be learned by practicing the invention; the main objectives and other advantages of this invention can be realized and obtained by means of the methods particularly pointed out in the description. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating the adaptive tension control method for loosely bonded prestressed beams based on a BP neural network provided in Embodiment 1 of the present invention. Figure 2 This is a schematic diagram of the cross-sectional structure and prestressing tendon arrangement of the rectangular cross-section beam in Embodiment 1 of the present invention; Figure 3 This is a graph showing the mean square error variation during the training process of the BP neural network in Embodiment 1 of the present invention. Detailed Implementation

[0021] The technical solutions of the present invention will be described in detail below through embodiments. The following embodiments are merely exemplary and can only be used to explain and illustrate the technical solutions of the present invention, and should not be construed as limiting the technical solutions of the present invention. Example

[0022] The geometric dimensions of the rectangular cross-section beam and the arrangement of the steel strands are as follows: Figure 2 As shown, the beam width is 600. 2500 28,000 meters long The diameter of a single steel strand is 15.2 mm. The cross-sectional area is 140 C40 concrete was used for pouring. Each beam consists of 24 steel strands, with each bonded steel strand weighing 1.35 kg. The length of the prestressed tendon is 13.808 mm. .

[0023] The beam reinforcement arrangement is as follows: Main reinforcement: 9 bars with a diameter of 32mm at the top of the cross-section The steel reinforcement consists of 20 reinforcing bars with a diameter of 32mm at the bottom of the cross-section. The uniformly distributed load contributed by this part is 1793.302. ; Torsional reinforcement: 28 bars, 14mm diameter The steel reinforcement contributes a uniformly distributed load of 332.024. ; Stirrups: 12mm in diameter Spacing 150 The limb clamps, with an in-interface length tentatively set at 12.8. 1 Based on 6.66 four-limb hoop calculations, the uniformly distributed load contributed by this part is: 741.862. ; Road surface load: Based on the material specifications, the estimated weight is approximately 1037.5 kg. The estimated area is 128.746405. Therefore, the total weight of the road surface is 133574.3951875. Converted to 28 On the beam, the uniformly distributed load contributed by this part is: 46751.0383 ; The self-weight of a beam includes: the weight of the prestressing tendons + the weight of the concrete + the weight of the reinforcing steel.

[0024] Upper construction load: The self-weight of the steel canopy, installed in the middle of the beam span, i.e., the column base of the steel canopy is in the middle of the beam span, and the column base cross-section is elliptical with a major axis length of 950 mm. Its major axis is aligned with the length of the beam, and its minor axis is 650 mm. .

[0025] To achieve the calculation of the tension control force for the aforementioned loosely bonded prestressed beams, the following technical solution is adopted: An adaptive tension control method for loosely bonded prestressed beams based on BP neural networks, such as... Figure 1 As shown, it includes the following steps: S1: Construct and train a BP neural network prediction model to establish the mapping relationship between the upper construction load, the stress at the bottom edge of the beam and the tension control force; S2: Obtain the structural parameters of the prestressed beam to be constructed and the current superstructure construction load information, and set the load control factor. Initial values, initial values ​​of upper construction loads, stress values ​​at the bottom edge of the beam, and increment step size. Among them, the stress value at the bottom edge of the beam is the limit of the concrete design tensile strength and the safety factor. The product; S3: Start the iterative optimization program. In each iteration, increment the step size. Cumulative update of load control factors Determine whether the load control factor is satisfied. If the condition is not met, proceed to step S4; if the condition is met, proceed to step S5. S4: Calculate the new upper construction load value as the product of the initial upper construction load value and the load control factor. Input the updated upper construction load value and the stress value at the bottom edge of the beam into the BP neural network prediction model, output the actual tension force under the current working condition, and return to step S3 to continue the iteration. S5: Order The process ends after inputting the initial value of the upper construction load and the stress at the bottom edge of the beam into the BP neural network prediction model and outputting the final actual tension control force. S6: Based on the actual tension force output in steps S4 and S5, calculate the edge stress at the bottom of the beam under the actual tension force. Determine edge stress Does it meet the crack resistance safety constraint conditions? ,in This is the standard value for the axial tensile strength of concrete; if it is not met, an alarm will be triggered.

[0026] In step S1, the training process of the BP neural network prediction model includes: S1.1: Construct a dataset containing the superstructure construction load, beam bottom edge stress, and tension control force. The tension control force is calculated from the elongation of the prestressing tendons at the tensioning end. The calculation process includes the friction loss of the loosely bonded prestressing tendons along the tensioning path to compensate for the tension control force. The formula for calculating the friction loss is as follows: In the formula, This represents the prestress loss value. For tension control force, The coefficient of influence of local deviation per meter of duct on friction ( ), To reduce the friction coefficient between the bonded prestressed steel bars and the sheath ( ), The length of the duct from the tensioning end to the calculated cross-section. The angle between the tensioning end and the tangent of the calculated section curve is the angle between the tensioning end and the tangent. The stress at the bottom edge of the beam is calculated using structural mechanics formulas or obtained through on-site measurement. S1.2: Normalize the data and construct a BP neural network with 1 hidden layer containing 7 neurons; the activation function between the input layer and the hidden layer is the Poslin positive linear transfer function, and the activation function between the hidden layer and the output layer is the Purelin linear transfer function. S1.3: Divide the dataset into training, validation and test sets with a ratio of 0.7:0.15:0.15; set the learning rate to 0.01, and use the upper construction load and the stress at the bottom edge of the beam as input values ​​and the tension control force as output values ​​for training; S1.4: Output the mean squared error curves of the training set, validation set, and test set during the training process of the BP neural network model, until the mean squared error is less than the preset accuracy threshold. Figure 3 As shown, the mean squared error (mse) of the 8th iteration reaches If the error is small enough and the prediction accuracy is high enough, the model is considered to have been successfully trained.

[0027] In step S2, the load control factor The method for determining the initial value is as follows: If it is the first stage of construction under the first level of load. The initial value is set to 0; if it is the subsequent [number]th [unit], then [the value is] 0. During the construction phase under high load, The formula for calculating the initial value is: ,in, This represents the cumulative load of the previous level. This is the cumulative design load for the current level.

[0028] In step S2, the incremental step size Δη is an adaptive adjustment value, and its calculation formulas are as follows: Formula ①, Formula ②, Formula ③, Formula ④, and Formula ⑤: ① ② ③ ④ ⑤ In the formula, ; The span and section height coefficients control the influence of structural flexibility. For the beam span, The height of the beam section. The performance coefficient of concrete controls the influence of concrete strength and stiffness. This refers to the standard value of the compressive strength of a concrete cube. The elastic modulus of concrete. This is the prestressing tendon ratio coefficient, used to control the influence of prestress reserve. This refers to the prestressed tendon reinforcement ratio. For stress level and safety reserve factor, The allowable tensile stress control value for the beam is determined according to the design specifications. This is the standard value for the axial tensile strength of concrete.

[0029] In step S6, the stress at the bottom edge of the beam under the actual tension is calculated. The calculation method is as follows:

[0030] In the formula: ( (total area of ​​steel strands) The bending moment generated by prestressing ( , (Eccentricity of prestressed ribs) The bending moment is the beam's self-weight plus the load from the superstructure. The cross-sectional area of ​​the beam is... The section modulus of the upper and lower edges, The effective stress of the prestressing tendon at section L is given.

[0031] Pressure sensors were installed in the upper construction area of ​​the beam to collect upper construction load data. Three to five sensors were installed in each span of the beam, evenly distributed in the span direction of the beam. Strain gauges were attached to the bottom center of the beam and the bottom of the beam side at the mid-span position to collect beam strain data. Three gauges were installed at each position. The sampling frequency was set to once every 12 hours.

[0032] In step S6, after the edge stress is detected to be not in line with the crack resistance safety constraint, an alarm action is automatically triggered within 1 second, and handling measures and alarm recording and tracing actions are performed. The alarm details are entered into the database and a ledger is formed.

[0033] The alarm actions include issuing audible and visual warning signals through an audible and visual alarm device and displaying them visually on the control terminal. The sound and light warning signals are issued in locations that cover the tensioning control terminal and the tensioning operation area at the construction site, within 5 meters of the tensioning end. The warning signals include flashing high-brightness LED warning lights and high-decibel buzzer sounds. The warning lights use a red constant-on and flashing mode, and the buzzer volume is not less than 85dB. The visualization is executed on the control terminal integrated with the tension control system. After the alarm is triggered, a red alarm pop-up window will automatically pop up, displaying the alarm type, alarm time, and current measured stress value.

[0034] Based on the actual working conditions of prestressed tensioning construction, and following the principle of "controlling risks first, then investigating and rectifying, and finally restoring," the measures are divided into three categories: emergency intervention measures, investigation and rectification measures, and construction restoration measures, to ensure that the handling of potential hazards is scientific, efficient, and safe.

[0035] Emergency intervention measures: Suspend tensioning operations, evacuate on-site personnel, and collect real-time monitoring data at high frequency; Inspection and rectification measures: Conduct inspections on excessive upper construction loads, prestressed tendon tension status, sensor data accuracy, and beam condition, and then implement rectification measures. Resumption of construction measures: Perform hazard verification, alarm reset, and issue a resumption of construction order.

[0036] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. An adaptive tension control method for loosely bonded prestressed beams based on a BP neural network, characterized in that, Includes the following steps: S1: Construct and train a BP neural network prediction model to establish the mapping relationship between the upper construction load, the stress at the bottom edge of the beam and the tension control force; S2: Obtain the structural parameters of the prestressed beam to be constructed and the current superstructure construction load information, and set the load control factor. Initial values, initial values ​​of upper construction loads, stress values ​​at the bottom edge of the beam, and increment step size. Among them, the stress value at the bottom edge of the beam is the limit of the concrete design tensile strength and the safety factor. The product; S3: Start the iterative optimization program. In each iteration, increment the step size. Cumulative update of load control factors Determine whether the load control factor is satisfied. If the condition is not met, proceed to step S4; if the condition is met, proceed to step S5. S4: Calculate the new upper construction load value as the product of the initial upper construction load value and the load control factor. Input the updated upper construction load value and the stress value at the bottom edge of the beam into the BP neural network prediction model, output the actual tension force under the current working condition, and return to step S3 to continue the iteration. S5: Order The process ends after inputting the initial value of the upper construction load and the stress at the bottom edge of the beam into the BP neural network prediction model and outputting the final actual tension control force. S6: Based on the actual tension force output in steps S4 and S5, calculate the edge stress at the bottom of the beam under the action of the actual tension force. Determine the edge stress Does it meet the crack resistance safety constraint conditions? ,in This is the standard value for the axial tensile strength of concrete; if it is not met, an alarm will be triggered.

2. The method according to claim 1, characterized in that, In step S1, the training process of the BP neural network prediction model includes: S1.1: Construct a dataset containing the superstructure construction load, the bottom edge stress of the beam, and the tension control force; the tension control force is calculated from the elongation value of the prestressing tendon at the tensioning end, wherein the calculation process takes into account the friction loss of the loosely bonded prestressing tendon on the tensioning path to compensate for the tension control force; the bottom edge stress of the beam is calculated by structural mechanics formulas or obtained by field measurement; S1.2: Normalize the data and construct a BP neural network with at least one hidden layer, wherein the hidden layer contains several neurons; the activation function between the input layer and the hidden layer adopts the Poslin positive linear transfer function, and the activation function between the hidden layer and the output layer adopts the Purelin linear transfer function. S1.3: Divide the dataset into a training set, a validation set, and a test set, set a fixed ratio and learning rate, and train the dataset using the upper construction load and the stress at the bottom edge of the beam as input values ​​and the tension control force as output values. S1.4: Output the mean squared error curves of the training set, validation set, and test set during the training process of the BP neural network model, until the mean squared error is less than the preset accuracy threshold, and the model is judged to have been successfully trained.

3. The method according to claim 1, characterized in that, In step S2, the load control factor The method for determining the initial value is as follows: If it is the first stage of construction under the first level of load. The initial value is set to 0; If it is the subsequent... During the construction phase under high load, The formula for calculating the initial value is: ,in, This represents the cumulative load of the previous level. This is the cumulative design load for the current level.

4. The method according to claim 1, characterized in that, In step S2, the incremental step size Δη is an adaptive adjustment value, and its calculation formula is as follows: Formula ①, Formula ②, Formula ③, Formula ④, and Formula ⑤: ① ② ③ ④ ⑤ In the formula, ; For span and section height coefficients, For the beam span, The height of the beam section. For concrete performance coefficients, This refers to the standard value of the compressive strength of a concrete cube. The elastic modulus of concrete. This is the prestressed tendon reinforcement ratio coefficient. This refers to the prestressed tendon reinforcement ratio. For stress level and safety reserve factor, This represents the allowable tensile stress control value for the beam. This is the standard value for the axial tensile strength of concrete.

5. The method according to claim 1, characterized in that, Pressure sensors are installed in the upper construction area of ​​the beam to collect upper construction load data. At least three sensors are installed in each span of the beam, evenly distributed in the span direction of the beam. Strain gauges are attached to the bottom center of the beam and the bottom of the beam side at the mid-span position to collect beam strain data. At least three strain gauges are installed at each position.

6. The method according to claim 1, characterized in that, In step S6, after the stress at the bottom edge of the beam is detected to be insufficient to meet the crack resistance safety constraint, an alarm is automatically triggered within 1 second. The alarm handling measures and alarm recording and tracing actions are then performed, and the alarm details are entered into the database and a ledger is formed.

7. The method according to claim 6, characterized in that, The alarm action includes issuing an audible and visual warning signal through an audible and visual alarm device and displaying it visually on the control terminal. The sound and light warning signals are emitted from locations covering the tensioning control terminal and the tensioning operation area at the construction site. The warning methods include flashing high-brightness LED warning lights and high-decibel buzzer sounds. The visualization is executed on the control terminal integrated with the tension control system. After the alarm is triggered, a red alarm pop-up window will automatically pop up, displaying the alarm type, alarm time, and current measured stress value.

8. The method according to claim 6, characterized in that, The processing measures include: Emergency intervention measures: Suspend tensioning operations, evacuate on-site personnel, and collect real-time monitoring data at high frequency; Inspection and rectification measures: Conduct inspections on excessive upper construction loads, prestressed tendon tension status, sensor data accuracy, and beam condition, and then implement rectification measures. Resumption of construction measures: Perform hazard verification, alarm reset, and issue a resumption of construction order.

Images