An automatic tracing and rectifying method for round curve steel box girder incremental launching construction
By using a real-time dynamic coordinate transformation model and a multi-dimensional attitude perception system, combined with a nonlinear adaptive correction algorithm, the problems of lag in manual correction and uneven alignment in the jacking construction of circular curve steel box girders were solved, achieving precise tracking and force balance of the beam and improving the level of intelligent construction.
Patent Information
- Application Number
- CN202611201003.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-10
- Publication Date
- 2026-11-03
AI Technical Summary
Existing technologies for launching curved steel box girders suffer from problems such as delayed manual correction, uneven alignment, and uneven lateral stress, making it difficult to achieve precise tracking of the girder's path and stress balance.
By employing a real-time dynamic coordinate transformation model and a multi-dimensional attitude perception system, combined with a nonlinear adaptive correction algorithm, and using a total station, inertial measurement unit, and hydraulic actuator, the beam's attitude is identified in real time and automatically corrected. A closed-loop control logic is constructed to ensure the beam's force balance and alignment accuracy during the jacking process.
It achieves millimeter-level automatic tracking control of the circular curve steel box girder, reduces frictional loss between the girder and the correction and limiting device, smooths attitude fluctuations during the jacking process, ensures the safe and accurate positioning of the large-span circular curve steel box girder in complex geographical environments, and improves the level of intelligence in bridge construction.
Smart Images

Figure CN123027798A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of engineering construction technology, specifically relating to an automatic tracking and correction method for the jacking construction of circular curve steel box girders. Background Technology
[0002] As modern transportation infrastructure construction continues to evolve towards greater complexity and precision, large-span steel box girders have become widely used in urban viaducts, overpasses, and large-span scenic bridges due to their significant advantages, such as light weight, high load-bearing capacity, high degree of prefabrication in factories, and short on-site assembly cycle. Among numerous construction techniques, the incremental launching method has become the mainstream solution for steel box girder erection in complex environments because it does not affect existing traffic below the bridge site, allows for the reuse of construction equipment, minimizes high-altitude work, and ensures controllable safety risks. Especially in engineering scenarios involving crossing existing railway trunk lines or busy highways, incremental launching has become almost the only choice to ensure the safe operation of existing lines and the successful completion of the bridge structure.
[0003] Within the existing framework of incremental launching construction technology, the walking-type incremental launching equipment is the core device for achieving beam displacement. Its working principle is typically based on the coordinated operation of longitudinal, transverse, and vertical hydraulic systems. Through a cyclical process of "lifting—pushing—lowering—returning," the massive steel box girder is driven forward step by step along a preset trajectory. During the incremental launching process on straight sections, the force model is relatively simple, and the control of the displacement trajectory mainly relies on the balanced distribution of longitudinal thrust. However, when the construction object transforms into a curved steel box girder, due to the constant curvature of the beam itself, its geometric characteristics determine that during the incremental launching process, the beam not only experiences tangential displacement along the design centerline but also inevitably experiences centripetal or centrifugal deviation tendencies due to geometric nonlinearity. To address this challenge, existing technologies typically employ physical constraints such as setting lateral guiding devices or limiting rollers on temporary piers, combined with periodic point measurements using a total station. When the measurement data indicates that the beam's axial deviation exceeds the allowable threshold, construction personnel manually adjust the thrust or stroke of the transverse jacks to forcibly correct the beam's attitude. This "monitoring-feedback-human intervention" model has, to some extent, solved the positioning problem of early low-curvature, short-distance jacking.
[0004] However, with the increasing demands for construction precision and intelligence in the engineering field, the aforementioned existing technical solutions have revealed a series of deep-seated technical limitations when dealing with the jacking of circular curves. The core contradiction lies in the imbalance between the dynamic continuity of the circular curve path and the discrete lag of manual correction. Specifically, during the movement of the circular curve beam, its ideal direction of travel at every instant deviates in real time with the displacement. Traditional manual measurement methods have significant time lags, often triggering correction actions only after the deviation has accumulated to a considerable extent. This delayed feedback mechanism easily leads to abrupt "step-like" displacements during the correction process, which not only makes it difficult to ensure the smoothness of the alignment but also generates huge, instantaneous lateral impact forces or frictional stresses between the beam and the limiting device, adversely affecting the precision coating of the steel box girder and the local stability of the structure.
[0005] Furthermore, the longitudinal thrust and lateral correction force in circular curve jacking have a complex dynamic coupling relationship. In existing technologies, lateral correction is often treated as an independent compensation action, failing to fully consider the combined interference of factors such as beam center of gravity shift, uneven friction coefficient, and temperature expansion and contraction on the tracking path during the jacking stroke. When the correction force is not properly controlled, it can easily cause torsional vibration or lateral instability of the beam at the support points, or even lead to shear failure of the support pad. In addition, traditional physical limiting devices are mostly rigid or semi-rigid designs. When facing the huge rotational inertia of large-span beams, they are not only costly to maintain, but also difficult to achieve micron-level fine attitude reshaping. The reason for this is the lack of a closed-loop control logic that can sense the spatial attitude of the beam in real time, automatically identify the deviation between the design line and the actual trajectory, and intelligently drive the hydraulic actuator to perform adaptive and coordinated correction.
[0006] Against this backdrop, how to achieve precise tracking of the beam's travel path under complex circular curve launching conditions, and how to construct an automatic correction mechanism that can dynamically mitigate deviations in real time and ensure structural stress balance, has become a key technical challenge and a pressing practical problem in the field of bridge engineering for improving the level of construction automation and ensuring the accuracy of alignment control for mega-bridges. Summary of the Invention
[0007] To address the technical problems mentioned in the background section regarding the lag, uneven alignment, and uneven lateral stress in manual correction during the jacking construction of circular curved steel box girders, this invention provides an automatic tracking and correction method for the jacking construction of circular curved steel box girders. This invention achieves precise identification of the circular curved girder's trajectory by constructing a real-time dynamic coordinate transformation model and a multi-dimensional attitude perception system. Furthermore, it combines a nonlinear adaptive correction algorithm to drive the hydraulic actuator, ensuring the girder's stress balance and alignment accuracy throughout the jacking process.
[0008] To achieve the above objectives, the present invention adopts the following technical solution: an automatic tracking and correction method for the jacking construction of circular curve steel box girders, comprising the following steps: The first step is to establish the mapping relationship between the global construction coordinate system and the local beam coordinate system: A total station observation reference network is set up at the starting end of the jacking of the circular curve steel box girder and at temporary piers along the line, and no fewer than four total reflection prisms are installed at specific locations on the steel box girder to be jacked; based on the design circular curve parameters, including the coordinates of the center (X... O Y O The theoretical trajectory mathematical model is constructed within the control host based on the curve radius R, the starting azimuth angle θ1, and the ending azimuth angle θ2. The coordinates (X, Y, F) of any point on the theoretical trajectory are given. t Y t ) satisfies the equation: (X t - X O ) 2 + (Y t - Y O ) 2 = R 2 ; The total station has an angle measurement accuracy of no less than 0.5 seconds and a distance measurement accuracy of no less than 0.5 mm + 1 ppm; the relative positional relationship between the center coordinates of the total reflection prism and the center line of the steel box girder structure is pre-calibrated by a coordinate measuring machine; the total station observation system is equipped with a meteorological compensation module, which collects temperature, air pressure and humidity data at the construction site in real time, and automatically corrects measurement errors caused by atmospheric refraction according to the atmospheric refraction correction model; The control host is equipped with a digital twin module, which records the thrust curve, displacement trajectory and correction response time in each jacking cycle in real time, and synchronously maps the spatial position of the steel box girder in the three-dimensional visualization interface; when abnormal lateral force fluctuations are detected at the support, the system automatically retrieves structural monitoring data and analyzes whether there is local support unevenness or slide plate damage, and generates maintenance suggestions. The total station observation reference network and the control host are equipped with a wireless data transmission link based on frequency hopping spread spectrum technology to ensure that the data transmission delay is controlled within 20 milliseconds. The control host integrates a fault diagnosis and emergency support module, which monitors the survival pulse of sensor data and hydraulic pipeline pressure. When an abnormality is detected, it automatically switches to a semi-automatic pressure holding mode or a short-time inertial tracking mode based on an inertial measurement unit. The theoretical trajectory mathematical model supports the extension of its application to complex linear steel beam launching projects with transition curve segments or variable curvature radii by adjusting the curvature parameter equations. The second step is to acquire the spatial pose data of the steel box girder in real time: This involves automatically tracking the total reflection prism using a total station to obtain its real-time three-dimensional coordinates (X, Y, F, Z) in the global construction coordinate system. i Yi Z i Meanwhile, the six-axis inertial measurement unit installed on the steel box girder senses the changes in the angular velocity and tilt attitude of the girder. The discrete prism coordinate data and the continuous inertial measurement data are filtered by a multi-sensor data fusion algorithm to output the real-time centroid coordinates and horizontal rotation orientation of each key position of the steel box girder. The sampling frequency of the total station is set to no less than 10Hz; the six-axis inertial measurement unit includes a three-axis accelerometer and a three-axis gyroscope, which are installed at the center of the transverse diaphragm of the steel box girder; the multi-sensor data fusion algorithm is the extended Kalman filter algorithm, which uses the low-frequency absolute coordinates obtained by the total station as the observed value and the high-frequency integral displacement obtained by the inertial measurement unit as the predicted value to correct the random noise generated by environmental construction vibration in real time; The third step is to calculate the deviation vector of the beam's trajectory: the control host calculates the deviation vector based on the real-time centroid coordinates (X...). s Y s ), calculate the normal distance from this point to the theoretical circular curve path as the real-time deviation Δd, the calculation formula is: By comparing the coordinate changes at adjacent sampling times, the angular deviation Δβ between the beam's tangential vector and the tangent direction of the theoretical circular curve is calculated; the deviation vector includes the deviation amount Δd and the deviation velocity V. d And the angular deviation △β; The direction of deviation is determined based on the sign of the deviation amount Δd: when Δd > 0, the beam is determined to be deviated to the outside of the curve; when Δd < 0, the beam is determined to be deviated to the inside of the curve; the system extracts the deviation vector as the input vector field of the correction control logic by analyzing the displacement vector changes within at least five consecutive sampling periods. The fourth step is to execute the dynamic correction logic based on the tracking algorithm: the correction algorithm based on fuzzy proportional integral differential control is adopted, and the output pressure and stroke of the transverse correction jacks symmetrically arranged on both sides of the bottom plate of the steel box girder are adjusted in real time according to the deviation vector. The pressure difference value ΔP of the hydraulic cylinders on both sides is calculated, and the correction torque is formed by adjusting the pressure difference to drive the beam to return to the theoretical trajectory. Before executing the correction action, the control unit first calculates the centripetal force compensation value during the circular curve jacking process; the control unit acquires the longitudinal thrust F in real time. z Calculate the longitudinal thrust F z The vector product of the current instantaneous radius of curvature is used to determine the reference lateral support force F. h0 The reference lateral support force F h0 The basic pressure holding value of the hydraulic system is applied to the lateral correction jack to counteract the centrifugal tendency of the beam moving along the tangent of the circular curve. The output logic of the correction algorithm is as follows: when the absolute value of Δd exceeds the preset first threshold δ1 and the angle deviation Δβ continues to increase, the host controller drives the lateral correction jack located in the opposite direction of the beam offset to increase the output pressure, while simultaneously reducing the pressure of the hydraulic cylinder in the same direction of offset; the pressure adjustment amount ΔP(t) is determined by the following formula: Where ΔP(t) represents the differential pressure output value required by the lateral correction jacks on both sides at time t, and its value is proportional to the correction torque; e(t) is the comprehensive weighted error of the deviation vector, which is synthesized from the deviation amount, deviation speed and angle deviation, and its weight coefficient is dynamically tuned according to the beam span and cantilever length; K p K i K d These are the proportional, integral, and differential gain coefficients, respectively, where the proportional gain K... p The integral gain K determines the sensitivity of the correction response to quickly suppress the current deviation. i The differential gain K is used to eliminate the long-term accumulated steady-state residual error of the system. d The prediction of the error change rate suppresses the oscillation caused by sudden disturbances. The three together constitute the core parameters of fuzzy proportional-integral-derivative control, and their values are dynamically adjusted according to the deviation state to adapt to the nonlinear working condition of circular curve pushing.
[0009] The correction algorithm includes a yaw angle damping compensation mechanism: when the inertial measurement unit detects a sudden increase in the horizontal yaw rate of the beam, the system applies a reverse damping torque through differential thrust to suppress the swaying tendency of the beam. The head of the lateral correction jack is equipped with a spherical crown structure with a displacement sensor to compensate for the longitudinal slope change of the steel box girder during the jacking process and the rotation caused by the lateral deformation of the girder; the feedback data of the displacement sensor and the macroscopic measurement data of the total station form a dual closed-loop control architecture, in which the total station is responsible for global position correction and the displacement sensor is responsible for the precise execution control of the correction stroke. The hydraulic actuator of the lateral correction jack adopts a proportional flow control valve with load-sensitive characteristics; at the initial stage of the correction action, mechanical clearance is eliminated by pre-pressurization with a small flow rate; during the correction process, the flow rate is dynamically adjusted according to the error change rate to ensure that the vector resultant direction of the correction speed and the longitudinal speed of the push is always along the tangent direction of the circular curve. The fifth step is multi-point collaborative synchronous control: Distributed controllers are deployed at each temporary support pier. The main control unit adopts a virtual spindle synchronous control strategy to decompose the total correction amount to each support point and control the transverse hydraulic cylinder at each support point to perform micro-displacement compensation during the longitudinal pushing stroke. The correction process and the longitudinal stroke of the step-by-step jacking are interlocked: when the longitudinal jacking displacement reaches the set step distance and the jacking force is stable, the lateral correction action enters a high-gain mode; when the beam is in the static jacking or return stage, the lateral correction mechanism maintains a pressure-holding state to resist external wind load interference; the system monitors the overall lateral displacement trend of the entire beam in real time by summing and averaging the lateral forces of all temporary pier supports in order to identify and filter out local environmental interference.
[0010] The automatic tracking and correction method also includes a path prediction step: the control host integrates a trajectory prediction model based on a deep neural network. This model takes the trajectory data of the historical jacking cycle, the current ambient temperature, and the structural stiffness matrix of the steel box girder as inputs to predict the possible deviation trend within the set stroke in the future. When the predicted potential deviation trend exceeds the warning value multiple times in a row, the system achieves feedforward pre-correction by adjusting the differential thrust of the longitudinal hydraulic cylinder. The theoretical trajectory mathematical model is dynamically corrected by introducing an elastic expansion compensation factor: the system dynamically corrects the theoretical radius R value in the control target based on the real-time collected average temperature of multiple points on the beam using the formula △R=α·R·△T, where α is the linear expansion coefficient of steel and △T is the temperature difference between the real-time temperature and the reference temperature, ensuring that the target line shape of the correction is consistent with the actual line shape of the structure after being affected by temperature. Friction monitoring units are installed at the support points of the steel box girder. These units consist of an array of pressure sensors embedded beneath a stainless steel sliding plate. The pressure sensors acquire the vertical reaction force R of the beam on the support points in real time. v The control host, combined with the real-time friction coefficient μ obtained through the initial jacking stage, verifies in real time whether the corrective force exceeds the lateral slip stability limit; if the proposed corrective force F... side >μ· R v The system triggers an overload alarm and automatically limits the rate of increase in correction pressure, gradually correcting the deviation by using multiple small compensations. For variable cross-section circular curve steel box girders, the control host adjusts the relative offset between the total reflection prism and the centerline in real time based on the pre-input structural CAD information to correct the tracking reference.
[0011] The correction algorithm includes a correction dead zone: when the deviation vector is within 30% of the design allowable error, the correction mechanism maintains the current pressure state; only when the deviation shows a continuous increasing trend or instantaneously exceeds the preset safety boundary, the system activates the high-efficiency correction mode.
[0012] This invention provides an automatic tracking and correction method for the jacking construction of circular curve steel box girders. Through the coordinated operation of the above-mentioned steps and the system, millimeter-level automatic tracking control of the complex trajectory of the circular curve is achieved. Compared with traditional technologies, this invention significantly reduces frictional losses between the girder and the correction and limiting device, smooths out attitude fluctuations during the jacking process, ensures the safe and accurate positioning of large-span circular curve steel box girders in complex geographical environments, and greatly improves the level of intelligence in bridge industrial construction.
[0013] The mathematical processing and logical judgments involved in this invention are all implemented by embedded software pre-installed in the main control computer. Its core algorithm, while meeting real-time requirements, possesses extremely high robustness. By combining the macroscopic positioning of the total station with the microscopic feedback of the displacement sensor, a complete closed loop of spatial attitude is constructed, fundamentally solving the control problem caused by geometric nonlinearity in circular curve jacking. Throughout the construction process, the system requires no frequent manual intervention; the correction actions are smooth and continuous, effectively avoiding damage to the beam structure caused by improper manual adjustments.
[0014] The key technical point of this invention lies in its dynamic compensation mechanism for centripetal force. By calculating in real time and applying lateral constraint forces that match the motion state, it transforms the original "passive collision" type of correction into an "active guidance" type of tracking. This transformation not only optimizes the stress distribution inside the steel box girder but also improves the adaptability of the jacking construction to different combinations of curvature radii and longitudinal and transverse slopes. Combined with the real-time capture of the dynamic vibration of the beam by the inertial measurement unit, the system can identify and suppress transient instabilities caused by hydraulic pulsation or support detachment, further ensuring construction safety.
[0015] The method described in this invention is not only applicable to circular curved steel box girders with a fixed curvature, but also, by adjusting the curvature parameter equations in the theoretical trajectory mathematical model, can be extended to the launching engineering of complex linear steel beams with transition curve segments or variable curvature radii. Its universal coordinate transformation logic and hierarchical control architecture possess strong industry applicability and technological scalability, representing an important development direction in the field of automated control for bridge construction.
[0016] In summary, this invention provides a scientific, rigorous, and reliable automated solution for the jacking construction of circular curved steel box girders by integrating multiple advanced technologies such as precision measurement, sensor fusion, adaptive control, and hydraulic servo drive. It not only solves the bottleneck of construction accuracy but also provides solid technical support for the rapid construction of long-span, complex curved bridges, possessing extremely high engineering application value and socio-economic benefits. Furthermore, this invention successfully constructs an intelligent control closed loop integrating perception, analysis, prediction, and execution through an automatic tracking and correction method for the jacking construction of circular curved steel box girders. This method not only solves the force balance problem of circular curved jacking at the physical level but also addresses the accuracy and continuity of trajectory control at the logical level, contributing an innovative technical path to enhancing the global competitiveness of my country's bridge construction technology. Attached Figure Description
[0017] Figure 1 This is a system integration architecture diagram of the automatic tracking and correction method of the present invention; Figure 2 This is a flowchart of the precision trajectory tracking and correction control of the present invention; Figure 3 This is a diagram of the environmental compensation and intelligent prediction control unit of the present invention. Detailed Implementation
[0018] Example 1
[0019] An automatic tracking and correction method for the incremental launching construction of circular curve steel box girders includes the following steps: The first step is to establish the mapping relationship between the global construction coordinate system and the local beam coordinate system: A total station observation reference network is set up at the starting end of the jacking of the circular curve steel box girder and at temporary piers along the line, and no fewer than four total reflection prisms are installed at specific locations on the steel box girder to be jacked; based on the design circular curve parameters, including the coordinates of the center (X... O Y O The theoretical trajectory mathematical model is constructed within the control host based on the curve radius R, the starting azimuth angle θ1, and the ending azimuth angle θ2. The coordinates (X, Y, F) of any point on the theoretical trajectory are given. t Y t ) satisfies the equation: (X t - X O ) 2 + (Y t - Y O ) 2 = R 2 ; like Figure 1The diagram shown is a system integration architecture diagram of the present invention. It illustrates the overall layout and data flow relationship of the total station observation reference network, control host, wireless data transmission link, and various hydraulic actuators. The total station has an angle measurement accuracy of no less than 0.5 seconds and a distance measurement accuracy of no less than 0.5 mm + 1 ppm. The relative position of the center coordinates of the total reflection prism and the center line of the steel box girder structure is pre-calibrated using a coordinate measuring machine. The total station observation system is equipped with a meteorological compensation module, which collects real-time temperature, air pressure, and humidity data from the construction site and automatically corrects measurement errors caused by atmospheric refraction based on an atmospheric refraction correction model. The control host is equipped with a digital twin module, which records the thrust curve, displacement trajectory and correction response time in each jacking cycle in real time, and synchronously maps the spatial position of the steel box girder in the three-dimensional visualization interface; when abnormal lateral force fluctuations are detected at the support, the system automatically retrieves structural monitoring data and analyzes whether there is local support unevenness or slide plate damage, and generates maintenance suggestions. The total station observation reference network and the control host are equipped with a wireless data transmission link based on frequency hopping spread spectrum technology to ensure that the data transmission delay is controlled within 20 milliseconds. The control host integrates a fault diagnosis and emergency support module, which monitors the survival pulse of sensor data and hydraulic pipeline pressure. When an abnormality is detected, it automatically switches to a semi-automatic pressure holding mode or a short-time inertial tracking mode based on an inertial measurement unit. The theoretical trajectory mathematical model supports the extension of its application to complex linear steel beam launching projects with transition curve segments or variable curvature radii by adjusting the curvature parameter equations. like Figure 1 The diagram shown is a system integration architecture diagram of the present invention, illustrating the overall layout and data flow relationship of the total station observation reference network, control host, wireless data transmission link and various hydraulic actuators.
[0020] The second step is to acquire the spatial pose data of the steel box girder in real time: This involves automatically tracking the total reflection prism using a total station to obtain its real-time three-dimensional coordinates (X, Y, F, Z) in the global construction coordinate system. i Y i Z i Meanwhile, the six-axis inertial measurement unit installed on the steel box girder senses the changes in the angular velocity and tilt attitude of the girder. The discrete prism coordinate data and the continuous inertial measurement data are filtered by a multi-sensor data fusion algorithm to output the real-time centroid coordinates and horizontal rotation orientation of each key position of the steel box girder. The sampling frequency of the total station is set to no less than 10Hz; the six-axis inertial measurement unit includes a three-axis accelerometer and a three-axis gyroscope, which are installed at the center of the transverse diaphragm of the steel box girder; the multi-sensor data fusion algorithm is the extended Kalman filter algorithm, which uses the low-frequency absolute coordinates obtained by the total station as the observed value and the high-frequency integral displacement obtained by the inertial measurement unit as the predicted value to correct the random noise generated by environmental construction vibration in real time; The third step is to calculate the deviation vector of the beam's trajectory: the control host calculates the deviation vector based on the real-time centroid coordinates (X...). s Y s ), calculate the normal distance from this point to the theoretical circular curve path as the real-time deviation Δd, the calculation formula is: By comparing the coordinate changes at adjacent sampling times, the angular deviation Δβ between the beam's tangential vector and the tangent direction of the theoretical circular curve is calculated; the deviation vector includes the deviation amount Δd and the deviation velocity V. d And the angular deviation △β; The direction of deviation is determined based on the sign of the deviation amount Δd: when Δd > 0, the beam is determined to be deviating outward from the curve; when Δd < 0, the beam is determined to be deviating inward from the curve. The system analyzes the displacement vector changes over at least five consecutive sampling periods and extracts the deviation vector as the input vector field for the correction control logic, such as... Figure 2 The diagram shown is a flowchart of the precision trajectory tracking and correction control of the present invention, which shows the complete logical link from coordinate mapping, pose acquisition, deviation vector calculation to dynamic correction output. The fourth step is to execute the dynamic correction logic based on the tracking algorithm: the correction algorithm based on fuzzy proportional integral differential control is adopted, and the output pressure and stroke of the transverse correction jacks symmetrically arranged on both sides of the bottom plate of the steel box girder are adjusted in real time according to the deviation vector. The pressure difference value ΔP of the hydraulic cylinders on both sides is calculated, and the correction torque is formed by adjusting the pressure difference to drive the beam to return to the theoretical trajectory. Before executing the correction action, the control unit first calculates the centripetal force compensation value during the circular curve jacking process; the control unit acquires the longitudinal thrust F in real time. z Calculate the longitudinal thrust F z The vector product of the current instantaneous radius of curvature is used to determine the reference lateral support force F. h0 The reference lateral support force F h0 The basic pressure holding value of the hydraulic system is applied to the lateral correction jack to counteract the centrifugal tendency of the beam moving along the tangent of the circular curve. The output logic of the correction algorithm is as follows: when the absolute value of Δd exceeds the preset first threshold δ1 and the angle deviation Δβ continues to increase, the host controller drives the lateral correction jack located in the opposite direction of the beam offset to increase the output pressure, while simultaneously reducing the pressure of the hydraulic cylinder in the same direction of offset; the pressure adjustment amount ΔP(t) is determined by the following formula: Where ΔP(t) represents the differential pressure output value required by the lateral correction jacks on both sides at time t, and its value is proportional to the correction torque; e(t) is the comprehensive weighted error of the deviation vector, which is synthesized from the deviation amount, deviation speed and angle deviation, and its weight coefficient is dynamically tuned according to the beam span and cantilever length; K p K i K d These are the proportional, integral, and differential gain coefficients, respectively, where the proportional gain K... p The sensitivity of the correction response is determined to quickly suppress the current deviation; the integral gain Ki is used to eliminate the long-term accumulated steady-state residual error of the system; and the differential gain K... d The prediction of the error change rate suppresses the oscillation caused by sudden disturbances. The three together constitute the core parameters of fuzzy proportional-integral-derivative control, and their values are dynamically adjusted according to the deviation state to adapt to the nonlinear working condition of circular curve pushing.
[0021] The correction algorithm includes a yaw angle damping compensation mechanism: when the inertial measurement unit detects a sudden increase in the horizontal yaw rate of the beam, the system applies a reverse damping torque through differential thrust to suppress the swaying tendency of the beam. The head of the lateral correction jack is equipped with a spherical crown structure with a displacement sensor to compensate for the longitudinal slope change of the steel box girder during the jacking process and the rotation caused by the lateral deformation of the girder; the feedback data of the displacement sensor and the macroscopic measurement data of the total station form a dual closed-loop control architecture, in which the total station is responsible for global position correction and the displacement sensor is responsible for the precise execution control of the correction stroke. The hydraulic actuator of the lateral correction jack adopts a proportional flow control valve with load-sensitive characteristics; at the initial stage of the correction action, mechanical clearance is eliminated by pre-pressurization with a small flow rate; during the correction process, the flow rate is dynamically adjusted according to the error change rate to ensure that the vector resultant direction of the correction speed and the longitudinal speed of the push is always along the tangent direction of the circular curve. The fifth step is multi-point collaborative synchronous control: Distributed controllers are deployed at each temporary support pier. The main control unit adopts a virtual spindle synchronous control strategy to decompose the total correction amount to each support point and control the transverse hydraulic cylinder at each support point to perform micro-displacement compensation during the longitudinal pushing stroke. The lateral correction process is interlocked with the longitudinal stroke of the step-by-step jacking: when the longitudinal jacking displacement reaches the set step distance and the jacking force is stable, the lateral correction action enters a high-gain mode; during the static jacking or return phase of the beam, the lateral correction mechanism maintains a pressure-holding state to resist external wind load interference; the system monitors the overall lateral displacement trend of the entire beam in real time by summing and averaging the lateral forces of all temporary pier supports to identify and filter out local environmental interference. Figure 3 The diagram shown is of the environmental compensation and intelligent prediction control unit of this invention, illustrating the collaborative working relationship between the meteorological compensation module, the temperature elastic stretching compensation factor, the deep neural network trajectory prediction model, and the friction monitoring unit.
[0022] The automatic tracking and correction method also includes a path prediction step: the control host integrates a trajectory prediction model based on a deep neural network. This model takes the trajectory data of the historical jacking cycle, the current ambient temperature, and the structural stiffness matrix of the steel box girder as inputs to predict the possible deviation trend within the set stroke in the future. When the predicted potential deviation trend exceeds the warning value multiple times in a row, the system achieves feedforward pre-correction by adjusting the differential thrust of the longitudinal hydraulic cylinder. The theoretical trajectory mathematical model is dynamically corrected by introducing an elastic expansion compensation factor: the system dynamically corrects the theoretical radius R value in the control target based on the real-time collected average temperature of multiple points on the beam using the formula △R=α·R·△T, where α is the linear expansion coefficient of steel and △T is the temperature difference between the real-time temperature and the reference temperature, ensuring that the target line shape of the correction is consistent with the actual line shape of the structure after being affected by temperature. Friction monitoring units are installed at the support points of the steel box girder. These units consist of an array of pressure sensors embedded beneath a stainless steel sliding plate. The pressure sensors acquire the vertical reaction force R of the beam on the support points in real time. v The control host, combined with the real-time friction coefficient μ obtained through the initial jacking stage, verifies in real time whether the corrective force exceeds the lateral slip stability limit; if the proposed corrective force F... side >μ· R v The system triggers an overload alarm and automatically limits the rate of increase in correction pressure, gradually correcting the deviation by using multiple small compensations. For variable cross-section circular curve steel box girders, the control host adjusts the relative offset between the total reflection prism and the centerline in real time based on the pre-input structural CAD information to correct the tracking reference.
[0023] The correction algorithm includes a correction dead zone: when the deviation vector is within 30% of the design allowable error, the correction mechanism maintains the current pressure state; only when the deviation shows a continuous increasing trend or instantaneously exceeds the preset safety boundary, the system activates the high-efficiency correction mode.
[0024] Example 2
[0025] In a large-span circular curved steel box girder launching project, the girder was 480 meters long with a designed curve radius R = 450 meters and a total launching weight of approximately 5200 tons. The automatic tracking and correction method described in this invention was employed during construction. The control system collected data from the total station and IMU every 100ms, and total reflection prisms were positioned at six locations: the beginning, end, and middle of the steel box girder. The lateral correction jack stroke was ±150mm, and the hydraulic system pressure control accuracy was 0.1MPa.
[0026] After the jacking operation entered the circular curve section, the system detected a tendency for the beam to shift outwards due to longitudinal thrust imbalance. At this point, Δd showed a shift of 4.2 mm, and Δβ yaw angle was 0.015 degrees. The system immediately triggered feedforward correction, calculated the centripetal force compensation pressure to be 3.5 MPa, and output differential pressure commands according to the Fuzzy-PID algorithm, causing the pressure of the outer correction cylinder to smoothly increase from 2.0 MPa to 5.8 MPa, while the pressure of the inner correction cylinder remained constant. In the subsequent 300 mm jacking stroke, the shift Δd rapidly converged to within 1.5 mm.
[0027] At the specific algorithm implementation level, the digital twin module described in this invention generates a three-dimensional visualization model in real time on the main control computer. This model not only maps the spatial coordinates of the beam but also displays the distribution of support reactions and lateral correction forces in real time through color cloud maps. When the value fed back by any sensor in the system deviates from the control limits defined by Statistical Process Control (SPC), the system automatically retrieves historical data streams for that time period for causal analysis. For example, when an abnormal increase in lateral force is detected at a specific pier location, the digital twin module combines the settlement monitoring data at that location to determine whether the lateral force redistribution of the beam is caused by uneven settlement of the temporary pier and immediately pushes maintenance suggestions to the engineering personnel.
[0028] The method of this invention exhibits strong adaptability when dealing with variable cross-section circular curved steel box girders. During implementation, if the width of the steel box girder varies linearly or curvilinearly along the longitudinal direction, the control host will read the pre-stored structural CAD parametric model in real time and dynamically correct the offset vector of the prism relative to the centerline based on the current jacking displacement L. This means that the control target point of the tracking algorithm always follows the design geometric center, rather than simply the physical edge, thus ensuring the positioning accuracy of the variable cross-section beam under complex alignment.
[0029] Furthermore, the load-sensitive proportional flow control technology employed in the hydraulic actuator sends a small pre-current signal to the proportional directional valve via the PID controller during the initial correction initiation. This causes the jack piston rod to extend slowly, eliminating physical gaps and elastic deformation between the mechanical components. Once the pressure feedback reaches the starting threshold, the flow regulation circuit immediately enters a high-gain mode, ensuring that the vector sum of the correction speed and the longitudinal jacking speed always points towards the tangent of the circular curve. This synergy of speed vectors eliminates the lateral shear stress between the beam and the slideway from a physical mechanism perspective, resulting in extremely smooth jacking operation.
[0030] To address the complex electromagnetic environment of construction sites, the wireless data transmission link involved in this invention employs Frequency Hopping Spread Spectrum (FHSS) technology and redundant error correction coding. During data exchange between the total station and the main control unit, at least 50 frequency switches are performed per second, ensuring that the transmission delay of control commands remains constant within 20ms even in strong interference environments near existing electrified railway bridges or high-voltage cables. This highly reliable communication guarantee is the physical basis for achieving millimeter-level dynamic correction, ensuring the determinism of the "sensing-computation-execution" control loop.
[0031] This invention also features a specially designed fault degradation handling process. During automatic tracking, the system continuously monitors the survival pulse of various sensors. If the total station loses its target due to heavy fog, the system will automatically and smoothly switch to "inertial tracking mode," using the IMU's integral displacement to maintain autonomous control for up to 300 seconds while simultaneously reducing the jacking speed. If the IMU data also becomes abnormal, the system triggers a secondary protection mechanism, automatically locking the pressure of all lateral hydraulic cylinders, switching the system to a semi-automatic pressure-holding state, and sending an alarm signal to the handheld terminal. This multi-level fault-tolerant design provides a solid safety foundation for the high-risk construction process of jacking curved steel box girders.
[0032] In optimizing the logic algorithm, this invention introduces a "dead zone" adjustment technique. Specifically, 30% of the allowable design error is defined as the insensitive zone. When the deviation vector is within this zone and shows no tendency to increase, the correction mechanism does not actively depressurize or pressurize, but only maintains the current static balance force. This strategy effectively solves the problem of frequent reciprocating switching of the hydraulic system within a very small error range, reduces the wear of the proportional valve, and makes the beam's trajectory more natural and smooth, conforming to the dynamic characteristics of large steel structures.
[0033] In summary, the automatic tracking and correction method for the jacking construction of circular curve steel box girders of this invention comprehensively solves the core pain points of difficult alignment control, complex stress, and strong manual lag in circular curve jacking by establishing precise coordinate mapping, fusing multi-source sensor data, applying adaptive correction control algorithms, and deploying a collaborative hydraulic execution system. Figure 1As shown in the diagram, the system integration architecture diagram of this invention intuitively illustrates the topological connection relationship between the total station observation reference network, the control host, and each hydraulic actuator; Figure 2 The flowchart of the precision trajectory tracking and correction control clearly presents the closed-loop control logic of "sensing-computation-execution"; Figure 3 The diagram of the environmental compensation and intelligent prediction control unit systematically illustrates the integrated architecture of multiple auxiliary modules, including meteorological compensation, temperature correction, trajectory prediction, and friction monitoring. This method not only achieves unmanned and precise intervention throughout the entire construction process but also, through cutting-edge technologies such as centripetal force compensation, path prediction, and temperature compensation, elevates the control precision and intelligence of bridge jacking construction in my country to new heights. Data comparison through examples demonstrates that this invention has significant industrial application value in reducing deviations, improving efficiency, and protecting structural safety, fully meeting the stringent requirements of modern transportation infrastructure construction for high precision and high safety.
Claims
1. An automatic tracking and correction method for the incremental launching construction of circular curve steel box girders, characterized in that, Includes the following steps: The first step is to establish the mapping relationship between the global construction coordinate system and the local beam coordinate system: A total station observation reference network is set up at the starting end of the jacking of the circular curve steel box girder and at temporary piers along the line, and no fewer than four total reflection prisms are installed at specific locations on the steel box girder to be jacked; based on the design circular curve parameters, including the coordinates of the center (X... O Y O The theoretical trajectory mathematical model is constructed within the control host based on the curve radius R, the starting azimuth angle θ1, and the ending azimuth angle θ2. The coordinates (X, Y, F) of any point on the theoretical trajectory are given. t Y t ) satisfies the equation: (X t - X O ) 2 + (Y t - Y O ) 2 = R 2 ; The second step is to acquire the spatial pose data of the steel box girder in real time: This involves automatically tracking the total reflection prism using a total station to obtain its real-time three-dimensional coordinates (X, Y, F, Z) in the global construction coordinate system. i Y i Z i Meanwhile, the six-axis inertial measurement unit installed on the steel box girder senses the changes in the angular velocity and tilt attitude of the girder. The discrete prism coordinate data and the continuous inertial measurement data are filtered by a multi-sensor data fusion algorithm to output the real-time centroid coordinates and horizontal rotation orientation of each key position of the steel box girder. The third step is to calculate the deviation vector of the beam's trajectory: the control host calculates the deviation vector based on the real-time centroid coordinates (X...). s Y s ), calculate the normal distance from this point to the theoretical circular curve path as the real-time deviation Δd, the calculation formula is: By comparing the coordinate changes at adjacent sampling times, the angular deviation Δβ between the beam's tangential vector and the tangent direction of the theoretical circular curve is calculated; the deviation vector includes the deviation amount Δd and the deviation velocity V. d And the angular deviation △β; The fourth step is to execute the dynamic correction logic based on the tracking algorithm: the correction algorithm based on fuzzy proportional integral differential control is adopted, and the output pressure and stroke of the transverse correction jacks symmetrically arranged on both sides of the bottom plate of the steel box girder are adjusted in real time according to the deviation vector. The pressure difference value ΔP of the hydraulic cylinders on both sides is calculated, and the correction torque is formed by adjusting the pressure difference to drive the beam to return to the theoretical trajectory. The fifth step is multi-point collaborative synchronous control: Distributed controllers are deployed at each of the temporary piers that participate in the support. The main control unit adopts a virtual spindle synchronous control strategy to decompose the total correction amount to each support point and control the transverse hydraulic cylinder at each support point to perform micro-displacement compensation during the longitudinal pushing stroke.
2. The automatic tracking and correction method for the jacking construction of a circular curve steel box girder according to claim 1, characterized in that, In the first step, the angle measurement accuracy of the total station is not less than 0.5 seconds, and the distance measurement accuracy is not less than 0.5mm + 1ppm; the relative positional relationship between the center coordinates of the total reflection prism and the center line of the steel box girder structure is pre-calibrated by a coordinate measuring machine; the total station observation system is equipped with a meteorological compensation module, which collects temperature, air pressure and humidity data of the construction site in real time, and automatically corrects the measurement error caused by atmospheric refraction according to the atmospheric refraction correction model; The control host is equipped with a digital twin module, which records the thrust curve, displacement trajectory and correction response time in each jacking cycle in real time, and synchronously maps the spatial position of the steel box girder in the three-dimensional visualization interface; when abnormal lateral force fluctuations are detected at the support, the system automatically retrieves structural monitoring data and analyzes whether there is local support unevenness or slide plate damage, and generates maintenance suggestions. The total station observation reference network and the control host are equipped with a wireless data transmission link based on frequency hopping spread spectrum technology to ensure that the data transmission delay is controlled within 20 milliseconds. The control host integrates a fault diagnosis and emergency support module, which monitors the survival pulse of sensor data and hydraulic pipeline pressure. When an abnormality is detected, it automatically switches to a semi-automatic pressure holding mode or a short-time inertial tracking mode based on an inertial measurement unit. The theoretical trajectory mathematical model supports the extension of its application to complex linear steel beam launching projects with transition curve segments or variable curvature radii by adjusting the curvature parameter equations.
3. The automatic tracking and correction method for the jacking construction of a circular curve steel box girder according to claim 1, characterized in that, In the second step, the sampling frequency of the total station is set to no less than 10Hz; the six-axis inertial measurement unit includes a three-axis accelerometer and a three-axis gyroscope, which are installed at the center of the transverse diaphragm of the steel box girder; the multi-sensor data fusion algorithm is the extended Kalman filter algorithm, which uses the low-frequency absolute coordinates obtained by the total station as the observed value and the high-frequency integral displacement obtained by the inertial measurement unit as the predicted value to correct the random noise generated by environmental construction vibration in real time.
4. The automatic tracking and correction method for the jacking construction of a circular curve steel box girder according to claim 1, characterized in that, In the third step, the direction of deviation is determined based on the sign of the deviation amount △d: when △d > 0, the beam is determined to be deviated to the outside of the curve; when △d < 0, the beam is determined to be deviated to the inside of the curve. The system extracts the deviation vector as the input vector field of the correction control logic by analyzing the displacement vector changes within at least five consecutive sampling periods.
5. The automatic tracking and correction method for the jacking construction of a circular curve steel box girder according to claim 1, characterized in that, In the fourth step, before performing the correction action, the control host first calculates the centripetal force compensation value during the circular curve jacking process; the control host acquires the longitudinal thrust F in real time. z Calculate the longitudinal thrust F z The vector product of the current instantaneous radius of curvature is used to determine the reference lateral support force F. h0 The reference lateral support force F h0 The basic pressure holding value of the hydraulic system is applied to the lateral correction jack to counteract the centrifugal tendency of the beam moving along the tangent of the circular curve. The output logic of the correction algorithm is as follows: when the absolute value of Δd exceeds the preset first threshold δ1 and the angle deviation Δβ continues to increase, the host controller drives the lateral correction jack located in the opposite direction of the beam offset to increase the output pressure, while simultaneously reducing the pressure of the hydraulic cylinder in the same direction of offset; the pressure adjustment amount ΔP(t) is determined by the following formula: Where ΔP(t) represents the differential pressure output value required by the lateral correction jacks on both sides at time t, and its value is proportional to the correction torque; e(t) is the comprehensive weighted error of the deviation vector, which is synthesized from the deviation amount, deviation speed and angle deviation, and its weight coefficient is dynamically tuned according to the beam span and cantilever length; K p K i K d These are the proportional, integral, and differential gain coefficients, respectively, where the proportional gain K... p The integral gain K determines the sensitivity of the correction response to quickly suppress the current deviation. i The differential gain K is used to eliminate the long-term accumulated steady-state residual error of the system. d The prediction of the error change rate suppresses the oscillation caused by sudden disturbances. The three together constitute the core parameters of fuzzy proportional-integral-derivative control, and their values are dynamically adjusted according to the deviation state to adapt to the nonlinear working condition of circular curve pushing. The correction algorithm includes a yaw angle damping compensation mechanism: when the inertial measurement unit detects a sudden increase in the horizontal yaw rate of the beam, the system applies a reverse damping torque through differential thrust to suppress the swaying tendency of the beam. The head of the lateral correction jack is equipped with a spherical crown structure with a displacement sensor to compensate for the longitudinal slope change of the steel box girder during the jacking process and the rotation caused by the lateral deformation of the girder; the feedback data of the displacement sensor and the macroscopic measurement data of the total station form a dual closed-loop control architecture, in which the total station is responsible for global position correction and the displacement sensor is responsible for the precise execution control of the correction stroke. The hydraulic actuator of the lateral correction jack adopts a proportional flow control valve with load-sensitive characteristics; at the initial stage of the correction action, mechanical clearance is eliminated by pre-pressurization with a small flow rate; during the correction process, the flow rate is dynamically adjusted according to the error change rate to ensure that the vector resultant direction of the correction speed and the longitudinal speed of the push is always along the tangent direction of the circular curve.
6. The automatic tracking and correction method for the incremental launching construction of a circular curve steel box girder according to claim 1, characterized in that, In the fifth step, the correction process and the longitudinal stroke of the step-by-step jacking are interlocked: when the longitudinal jacking displacement reaches the set step distance and the jacking force is stable, the lateral correction action enters a high-gain mode; when the beam is in the static jacking or return stage, the lateral correction mechanism maintains a pressure-holding state to resist external wind load interference; the system monitors the overall lateral displacement trend of the entire beam in real time by summing the lateral forces of all temporary pier supports and taking the average value, so as to identify and filter out local environmental interference.
7. An automatic tracking and correction method for the jacking construction of a circular curve steel box girder according to any one of claims 1 to 6, characterized in that, The automatic tracking and correction method also includes a path prediction step: the control host integrates a trajectory prediction model based on a deep neural network. This model takes the trajectory data of the historical jacking cycle, the current ambient temperature, and the structural stiffness matrix of the steel box girder as inputs to predict the possible deviation trend within the set stroke in the future. When the predicted potential deviation trend exceeds the warning value multiple times in a row, the system achieves feedforward pre-correction by adjusting the differential thrust of the longitudinal hydraulic cylinder. The theoretical trajectory mathematical model is dynamically corrected by introducing an elastic expansion compensation factor: the system dynamically corrects the theoretical radius R value in the control target based on the real-time collected average temperature of multiple points on the beam using the formula △R=α·R·△T, where α is the linear expansion coefficient of steel and △T is the temperature difference between the real-time temperature and the reference temperature, ensuring that the target line shape of the correction is consistent with the actual line shape of the structure after being affected by temperature. Friction monitoring units are installed at the support points of the steel box girder. These units consist of an array of pressure sensors embedded beneath a stainless steel sliding plate. The pressure sensors acquire the vertical reaction force R of the beam on the support points in real time. v The control host, combined with the real-time friction coefficient μ obtained through the initial jacking stage, verifies in real time whether the corrective force exceeds the lateral slip stability limit; if the proposed corrective force F... side >μ· R v The system triggers an overload alarm and automatically limits the rate of increase in correction pressure, gradually correcting the deviation by using multiple small compensations. For variable cross-section circular curve steel box girders, the control host adjusts the relative offset between the total reflection prism and the centerline in real time based on the pre-input structural CAD information to correct the tracking reference.
8. An automatic tracking and correction method for the jacking construction of a circular curve steel box girder according to any one of claims 1 to 6, characterized in that, The correction algorithm includes a correction dead zone: when the deviation vector is within 30% of the design allowable error, the correction mechanism maintains the current pressure state; only when the deviation shows a continuous increasing trend or instantaneously exceeds the preset safety boundary, the system activates the high-efficiency correction mode.
