A control method for rapid matching and assembling of large-span steel-concrete composite beam sections

Abstract

The application discloses a kind of large-span steel and concrete composite beam section fast matching assembly control methods, it is related to bridge technical field, the method is by obtaining beam section end face humidity data and original beam body state data, analysis end face distortion displacement and rotation state, form predeformation result;In the process of cantilever erection, combined with beam section rotation state and predeformation result to determine the positioning mode, and identify contact interface friction state parameters and the displacement coupling relationship between support leg, to carry out collaborative adjustment to each support leg, realize beam section end face gradually accurate alignment;In the alignment process, according to the stress and beam body strain data of support leg are executed stress coordination processing, complete section unloading and form reset lock or repositioning assembly branch mechanism;After alignment is completed, end face connecting bolt is executed screwing or welding processing, and records the assembly datum of subsequent beam section, the method is suitable for high-precision assembly of section in the process of bridge cantilever erection construction.

Description

Technical Field

[0001] This invention relates to the field of bridge technology, specifically to a method for rapid matching and assembly control of long-span steel-concrete composite beam segments. Background Technology

[0002] Long-span steel-concrete composite bridges are widely used in cantilever construction of overpasses, suspension bridges, and cable-stayed bridges due to their combination of the lightweight and high strength of steel structures and the compressive strength of concrete. The on-site cantilever erection of these bridges typically employs segmental prefabrication, cantilever assembly, or segmental erection by a bridge-building machine. The accuracy of the matching and alignment of the beam segment ends directly determines the overall bridge alignment and closure quality.

[0003] Existing segmental erection and assembly control methods mostly rely on manual measurement with total stations and manual adjustment with jacks, which has the following shortcomings: First, existing technologies generally use rigid body kinematic models, ignoring the force and position coupling effect between the legs caused by the flexible deformation of the beam. During the erection process, the lifting of each leg interferes with each other, and the actual position deviates significantly from the command. Furthermore, repeated adjustments cause the friction state of the support contact surface to evolve in a path-dependent manner, resulting in a mismatch in actual response, thus causing low erection efficiency and difficulty in meeting the closure requirements. Second, during the prefabrication storage and assembly waiting period, the end face of the steel-concrete composite bridge segment undergoes non-uniform shrinkage distortion due to humidity gradient, causing the end face geometry to deviate from the design plane. Forced alignment during erection will form a ring-shaped hidden gap, and the rapid erection process will cause the gap to be locked by stress, seriously affecting the durability and long-term safety of the bridge deck. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a rapid matching and assembly control method for large-span steel-concrete composite beam segments, solving the problems in the background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for rapid matching and assembly control of segments in a large-span steel-concrete composite beam includes: Obtain humidity data and original beam condition data of the end face of the steel-concrete composite beam segment to be erected, determine the distortion displacement and rotation state of the beam segment end face, and generate pre-deformation results; During the cantilever erection of the bridge, the rotation state of the beam segment is identified and updated. Combined with the pre-deformation results, the triggering of the adjustment command is determined, and the friction state parameters of the contact interface are identified. The displacement coupling relationship between the legs is identified by probing the displacement. Based on the adjustment command and friction state parameters, the position of each outrigger is adjusted to achieve precise alignment of the beam segment end face. The adjustment includes the command stroke and erection speed state of the outrigger coordinated operation. During the erection and alignment process, stress coordination processing is performed based on the force data of the outriggers and the strain data of the beam to complete the unloading of the cantilever erection segment and re-determine the triggering of the adjustment command to form a branch assembly mechanism, including reset locking and readjustment assembly. After the erection and alignment are completed, the connecting bolts at the end face of the current steel-concrete composite beam segment are tightened or welded, and the assembly reference of the next steel-concrete composite beam segment to be erected is recorded.

[0006] The above-described solution of the present invention has at least the following beneficial effects: By introducing an assembly benchmark that combines end-face distortion displacement with pre-deformation results during cantilever erection, deviations caused by concrete shrinkage can be considered simultaneously during the alignment stage of the beam segments. This results in an end-face fit path that matches the bridge alignment control requirements during actual erection. At the same time, by continuously updating the rotation state of the beam segments and dynamically correcting the assembly benchmark, the alignment of each segment remains continuous during spatial connection, reducing the cumulative deviation caused by cantilever construction. This ensures that the overall bridge structure maintains a stable shape and stress transmission relationship during segment-by-segment erection.

[0007] By coordinating the position adjustment of the outriggers under multi-leg support conditions and performing stress coordination treatment in combination with the stress state of the outriggers and the strain response of the beam, a balanced stress distribution path is formed in the beam segment during alignment and unloading. At the same time, during the end-face connection stage, differentiated pre-tightening is applied to the bolts or welded parts according to the actual gap distribution, so that the connection interface gradually transitions from the initial contact state to a structural connection form with a consistent overall stress state, thereby reducing the cumulative impact of local stress concentration and structural incoordination during the bridge segment assembly process. Attached Figure Description

[0008] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0009] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0010] like Figure 1 As shown, an embodiment of the present invention provides a method for rapid matching and assembly control of large-span steel-concrete composite beam segments, including: S100: Obtain the humidity data of the end face of the steel-concrete composite beam segment to be erected and the original beam state data, determine the distortion displacement and rotation state of the beam segment end face, and generate the pre-deformation results; S200: During the cantilever erection of the bridge, the rotation state of the beam segment is identified and updated. Combined with the pre-deformation results, the triggering of the adjustment command is determined, and the friction state parameters of the contact interface are identified. The displacement coupling relationship between the legs is identified by probing the displacement. S300: Based on the adjustment command and friction state parameters, adjust the position of each outrigger to achieve precise alignment of the beam segment end face. The adjustment includes the command stroke and erection speed state of the outrigger coordinated operation. S400: During the erection and alignment process, stress coordination processing is performed based on the force data of the outriggers and the strain data of the beam to complete the unloading of the cantilever erection segment and re-determine the triggering of the adjustment command to form a branch assembly mechanism, including reset locking and readjustment assembly. S500: After the erection and alignment are completed, tighten or weld the connecting bolts at the end face of the current steel-concrete composite beam segment, and record the assembly reference of the next steel-concrete composite beam segment to be erected.

[0011] In this embodiment of the invention, the accuracy and adaptability of cantilever erection are improved by implementing closed-loop control throughout the entire process, from monitoring the humidity of the concrete end face to final locking compensation. Specifically, the method first uses humidity data to predict the dish-shaped distortion of the concrete end face caused by the humidity gradient, and actively generates a reverse pre-deformation benchmark. This is equivalent to anticipating and offsetting the future shrinkage deformation of the end face before assembly, avoiding the hidden gaps formed by forced alignment in traditional methods. Subsequently, during the repositioning process, the path-dependent evolution of the friction state of the support pads and the flexible coupling effect between the outriggers are identified in real time, and the control model is corrected online. This ensures that repeated repositioning will not result in the implicit accumulation of deviations. For example, during the summer construction of a cross-sea bridge, the humidity of the end face changes rapidly due to sunlight and the support pads slide repeatedly. Traditional methods often require several trial adjustments to barely close the bridge, while this method reduces the number of trial adjustments by using friction parameter feedforward and adaptive update of the flexibility matrix. Meanwhile, stress coordination control and micro-unloading mechanism ensure that the beam is subjected to uniform stress during the erection process. Once the reaction force or strain of a certain leg exceeds the limit, all legs will simultaneously decrease slightly to redistribute the stress and prevent local overload damage. Finally, in the connection stage, non-uniform pre-tightening force is distributed according to the measured end face gap, and reverse pre-deformation is applied to the welding thermal shrinkage so that the beam segment line automatically returns to the correct shape after cooling.

[0012] This embodiment solves the problem of pose instability and implicit accumulation of deviations caused by end face deformation, support path dependence and welding delay in cantilever erection by constructing a four-dimensional adaptive inverse control mode based on humidity distortion, friction evolution, flexible coupling and thermal shrinkage compensation. This enables the system to accurately identify the compliance convergence fingerprint characteristics under complex process conditions from the mechanical topology, and realize the rapid and accurate locking of large-span steel-concrete composite beam segments.

[0013] In a preferred embodiment of the present invention, S101: In the bridge material area, such as the concrete area, at both ends of the steel-concrete composite beam segment to be erected, micro-holes are pre-drilled according to a spatial orthogonal grid layout. Different hole depths can be set, for example, 0.2 times, 0.5 times, and 0.8 times the end face thickness, respectively. At least a number of humidity sensor probes are arranged at equal intervals in the transverse and vertical directions along the end face at each depth level. The sensor probes are selected from capacitive polymer humidity sensitive elements, with a measurement accuracy of ±1.5%RH and a response time of no more than 30 seconds. By recording the relative humidity changes at each measuring point on the end face of the steel-concrete composite beam segment to be erected, humidity data is obtained to provide initial boundary conditions for subsequent humidity diffusion prediction. S102: Based on humidity data, the hydraulic control system calculates the current humidity gradient distribution in real time, and analyzes the humidity diffusion state of the concrete in the bridge material through Fick's second law. It predicts the uneven shrinkage trend of the end face during the future assembly window before the beam segment is stored and cantilevered, and obtains the distortion displacement value at each position of the end face. In practice, humidity data is input into an edge computing server. The server pre-stores material parameters for the beam segment, including the functional relationship between the moisture diffusion coefficient and the water-cement ratio and aggregate gradation, the ultimate shrinkage strain value, and the shrinkage rate coefficient. Based on the unsteady-state humidity diffusion control equation, i.e., Fick's second law, the end face is discretized into spatial grid cells using the finite difference method. The time step is determined according to the computational convergence requirements, and the boundary condition is set as the relative humidity of the exposed environment of the end face, provided in real-time by the on-site weather station. By iteratively solving this equation, the humidity distribution inside the end face at any point between the current moment and the expected assembly time is predicted. Because the humidity varies gradient along the thickness direction of the end face (e.g., lower at the surface and higher inside), the shrinkage amount differs at different depths, causing bending deformation of the end face.

[0014] Therefore, an exponential function is used to predict the non-uniform shrinkage trend of the end face during the future assembly window. Specifically, first, the difference between the predicted humidity and the initial humidity is calculated, and then the difference between the predicted humidity and the saturation state is calculated. The ratio of the two differences is used as an indicator of the degree of humidity change, reflecting the degree of decline of the current humidity relative to the saturation state. The larger the value, the greater the decrease in humidity and the greater the shrinkage driving force. Then, the ratio is multiplied by the shrinkage rate coefficient and an exponential operation is performed to make the shrinkage strain increase non-linearly with the decrease in humidity. Then, the exponential value is subtracted from 1, and finally multiplied by the limit shrinkage strain value to obtain the free shrinkage strain value at the corresponding time at that location. This calculation process traverses all discrete points of the end face to obtain the non-uniform shrinkage strain distribution of each depth layer of the entire end face.

[0015] Free shrinkage strain reflects the degree of volume shrinkage caused by the evaporation of moisture and decrease in humidity within the concrete under conditions without any external constraints.

[0016] The free contraction strain of each depth layer is converted into equivalent mechanical load by the equivalent temperature load method. For example, the equivalent temperature load method is used and input into the end face finite element elastic model. The finite element solver is run to calculate the thermal expansion or contraction deformation of the end face under the action of the equivalent temperature load. The self-generated deformation displacement field of the end face under the condition of no external constraint is solved. This displacement field describes the distortion displacement value of the end face from the initial planar state to the dish or warped shape. This displacement value reflects the degree of distortion of the end face geometry. The larger the value, the more serious the deformation of the end face caused by humidity contraction. If it is not compensated during assembly, it will lead to uneven gaps between the end face and the erected beam segment, that is, the center is pressed and the edges are opened, forming hidden gaps and dish distortion. The distortion displacement value is the displacement component perpendicular to the end face direction. A positive value indicates that the point is convex relative to the theoretical plane, and a negative value indicates that it is concave.

[0017] By pre-measuring or predicting the distortion displacement value, the hydraulic system can compensate in reverse, aligning the end face with the pre-deformation result, thereby offsetting the effects of future shrinkage distortion.

[0018] Initial humidity refers to the internal relative humidity of concrete after curing and before drying shrinkage, serving as the starting point for humidity changes; saturation represents a relative humidity of 100%RH, i.e., completely wet; the shrinkage rate coefficient is a material constant used to reflect the rate of concrete shrinkage. Subtracting the exponent value from 1 reflects that when the humidity has not yet decreased, the ratio is 0, the exponent value is 1, and the shrinkage strain is 0; when the humidity decreases and approaches saturation, or when it is dry for a long time, the exponent value approaches 0, and the shrinkage strain approaches the limit shrinkage strain value. S103: Based on the distortion displacement value, the reverse compensation amount corresponding to each outrigger is set in the hydraulic control system to identify the pre-deformation result, and the drive end face is aligned with the pre-deformation result during cantilever erection, serving as the assembly target for subsequent posture control.

[0019] The distortion displacement value predicted by S102 is vector-superimposed with the theoretical design datum. The distortion displacement value describes the convex or concave shape of the end face caused by the humidity gradient. The distortion displacement value is then superimposed in reverse, i.e., the distortion component is subtracted from the theoretical design datum, to obtain the pre-deformation result, i.e., the pre-deformation datum. This result is used to describe that if the beam segment end face is assembled and aligned according to the shape of the pre-deformation datum, then when the end face undergoes actual humidity shrinkage distortion in the future, it will deform exactly into the theoretical design shape. It is updated in real time as the assembly time approaches. The closer to the assembly time, the more accurate the predicted distortion, and the closer the pre-deformation datum is to the final target.

[0020] Next, the controller sends the reverse compensation amount of each outrigger to the servo valve in the form of an electrical signal. The servo valve adjusts the flow of hydraulic oil into the outrigger cylinder, driving the piston rods of each outrigger to move synchronously, causing the end face of the beam segment to gradually deviate from the theoretical design plane to achieve the pre-deformation result. This pre-deformation result serves as the assembly target for subsequent pose control. That is, subsequent pose deviation calculations and adjustment command generation are all based on this pre-deformation result, rather than the theoretical design plane, thus ensuring that after the end face undergoes humidity shrinkage distortion in the future, its actual shape will exactly return to the theoretical design position.

[0021] The reverse compensation amount reflects the reverse displacement that the outrigger needs to actively generate at the current moment in order to counteract the humidity shrinkage distortion that will occur on the end face in the future; that is, the opposite of the distortion displacement value. In this embodiment of the invention, the non-uniform shrinkage of concrete material caused by humidity gradients is transformed from an uncontrollable distortion error into a predictable reverse benchmark. An embedded humidity sensor array is used to capture the humidity evolution at different depths of the end face in real time. Fick's second law and a finite element model are used to quantitatively predict the dish-shaped distortion displacement field of the end face during the future assembly window. This distortion field is then superimposed onto the theoretical design plane to generate a pre-deformation benchmark that updates dynamically over time. This strategy allows the hydraulic system to no longer target a static ideal plane, but instead actively drive the end face to present a pre-deformation posture opposite to the future shrinkage trend, thereby transforming the end face's own deformation from a source of assembly deviation into a pre-compensable amount.

[0022] This completely eliminates the hidden gaps and false deviation signals caused by end-face humidity distortion, allowing subsequent pose comparison and adjustment commands to directly reflect the real installation error rather than end-face deformation artifacts. This further reduces the number of iterations and the risk of misadjustment in closed-loop adjustment, thereby ensuring the initial reference accuracy and convergence speed of end-face matching during cantilever erection.

[0023] In a preferred embodiment of the present invention, S201: At five feature points—the four corner points and the center of the end face of the beam segment to be assembled—integrated sensor modules of a triaxial microelectromechanical system accelerometer and a biaxial electrolyte inclinometer are installed, respectively. Each module can output three translational acceleration signals and two tilt angle signals for that point. A laser displacement sensor reflective target is arranged at several-meter intervals along the length of the outer surface of the web on both sides of the beam segment. Simultaneously, laser displacement sensor transmitters are installed at corresponding positions on the already erected beam segments. The initial distance between the transmitter and the target is calibrated using a measuring tape. A set of magnetostrictive displacement sensors is installed at the lifting points of the four legs at the bottom of the beam segment to measure the relative displacement of the legs relative to the bottom surface of the beam segment in real time.

[0024] The four corner points reflect the overall tilt and torsion of the beam segment in the plane; the end face center reflects the relative deformation of the end face center point with respect to the four corner points, which is used to detect whether the end face has local warping. Obtain the original beam state data of the steel-concrete composite beam segment to be erected, and analyze the rotation state of the steel-concrete composite beam segment to be erected based on the original beam state data to obtain the beam segment rotation data. The original beam state data includes the relative displacement values, biaxial tilt angle values, and triaxial acceleration values ​​of each measuring point, which are used to reflect the local motion state of the beam segment. The biaxial tilt angle value includes two angular components output by each measuring point: the pitch angle about the transverse bridge axis and the roll angle about the longitudinal bridge axis, which directly reflects the local attitude at the measuring point. The triaxial acceleration value refers to the linear acceleration along the three orthogonal directions of X, Y, and Z measured by the triaxial microelectromechanical system accelerometer at each measuring point, which is used to help determine the dynamic vibration or inertial motion of the beam segment. The biaxial tilt angle value is obtained through a biaxial electrolyte tilt meter, and the triaxial acceleration value is obtained through a triaxial microelectromechanical system accelerometer. These two sensors are integrated into the same module and installed at the four corner points and end face center of the beam segment to be assembled. They are used to measure the pitch angle and roll angle of the beam segment relative to the horizontal plane, as well as the linear acceleration in three orthogonal directions.

[0025] The beam segment rotation data includes the pitch angle of the beam segment around the transverse bridge axis, the roll angle around the longitudinal bridge axis, the yaw angle around the vertical axis, and the translational degrees of freedom in the X, Y, and Z directions. This data is used to calculate the deviation between the current position and orientation of the beam segment in space and the target, thereby driving the hydraulic system to generate adjustment commands so that the beam segment gradually approaches the pre-deformation result.

[0026] Using the tilt angle data of three non-collinear points, an overdetermined plane is used. By filtering out small local deformation errors through the least squares method, the best fitting plane of the beam segment as a rigid body in the global coordinate system is calculated, thereby accurately obtaining the overall pitch angle and roll angle. The difference between the real-time readings of the four outrigger displacement sensors and the initial calibration values ​​is used, combined with geometric relationships, to solve the overall vertical translation and yaw angle of the beam segment about the vertical axis. Specifically, when calculating the overall rotation angle of the beam segment around the transverse and longitudinal bridge axes based on the inclination angle values ​​of three non-collinear feature points, the system first reads the two inclination angles output by the biaxial inclinometer at each feature point, corresponding to the pitch angle around the transverse bridge axis and the roll angle around the longitudinal bridge axis, respectively. Since the beam segment may experience local elastic deformation during the repositioning process, the measured inclination angle values ​​at each point are not entirely consistent. To obtain the overall rigid body rotation angle of the beam segment, the system uses the least squares method for fitting: the pitch angle measurements of all feature points are used as observation data, assuming the overall pitch angle of the beam segment is a constant to be determined, the sum of squares of the deviations of each measurement value from this constant is calculated, and the constant that minimizes this sum of squares is the optimal estimate of the overall pitch angle; similarly, the roll angle is processed in the same way. This method can effectively suppress the interference of local deformation or measurement noise on the overall attitude calculation.

[0027] For the overall vertical translation and yaw angle of the beam segment, magnetostrictive displacement sensors at the lifting points of the four outriggers are used. In the initial calibration state of the beam segment, typically when it has just been hoisted into place and has not yet been adjusted, the initial displacement value of each outrigger is recorded. During the adjustment process, the current displacement value of each outrigger is read in real time, and the difference between the current displacement and the initial calibration value of each outrigger is calculated. This difference is the relative lifting amount at that outrigger. The arithmetic mean of the lifting amounts of all four outriggers is then used to obtain the overall vertical translation of the beam segment.

[0028] For the yaw angle about the vertical axis, select two symmetrical outriggers on the left and right sides, calculate the difference in their lift, and then divide it by the lateral distance between the two outriggers. The arctangent of the resulting ratio is the yaw angle. This method is based on the small-angle approximation in rigid body kinematics. When the yaw angle is small, the difference in lift can be directly divided by the lateral distance to obtain the radian value.

[0029] S202: Install Beidou reference stations at the base of bridge piers or towers to monitor the rotational displacement of the erected beam segments in real time during cantilever erection. Correct the measurement reference through reference conversion, thereby eliminating errors in the measurement results caused by the rotation or translation of the erected beam segments and updating the beam segment rotation data. This facilitates the correction of measurement reference drift caused by cantilever erection deformation and ensures the stability of the assembly reference during cantilever erection.

[0030] The erected beam segments refer to those bridge segments that have been assembled, locked, and supported on the piers or cantilever ends before the current beam segments to be assembled.

[0031] Because the erected beam segments will slowly displace and rotate under the influence of subsequent segmental loads, cable tension, and concrete creep, the reference point of the laser displacement sensor installed on the erected beam segments will drift. To eliminate this error, three BeiDou reference stations are deployed in stable areas at the top of the bridge piers and the base of the towers, unaffected by construction deformation. Each reference station is equipped with a high-precision choke coil antenna, capable of outputting centimeter-level absolute coordinates in real time.

[0032] Simultaneously, prisms are installed at multiple control points of the erected beam segments, and the absolute coordinate changes of these prisms are periodically measured using a total station. These changes are input into a reference drift compensation filter to calculate the measurement reference drift components of the erected beam segments from their initial state to their current state, including the rotation matrix and translation vector. Since the rotational drift of the erected beam segments during construction is relatively small, satisfying the small-angle approximation condition, the calculated beam segment rotation data can be directly subtracted from the measurement reference drift components to update the beam segment rotation data, thus eliminating the influence of reference drift. S203: During the cantilever erection process, the updated beam segment rotation data is compared with the pre-deformation results. If the comparison result exceeds the preset comparison threshold, it indicates that the deviation does not meet the fine assembly requirements. In this case, the adjustment command is triggered to start the hydraulic adjustment. The comparison result is an instantaneous deviation vector obtained by calculating the difference between the updated beam segment rotation data and the pre-deformation result. It includes the deviation between various features. If the absolute value of each component of the deviation vector is less than the corresponding threshold, it is determined that the current beam segment pose has met the fine assembly requirements. The system skips the subsequent adjustment steps and directly enters the locking process. If any component exceeds the threshold, it is determined that adjustment is required. The system records the current deviation vector and generates an adjustment command.

[0033] In this embodiment of the invention, the invention eliminates the reliance on laser point cloud or image processing, and directly calculates the six-degree-of-freedom pose of the beam segment based on acceleration, tilt angle and displacement signals, thereby improving the measurement robustness and environmental adaptability in cantilever erection.

[0034] On the one hand, inertial and displacement sensors deployed at characteristic points of the beam segment are used to capture the spatial motion state of the beam segment in real time from a physical perspective, avoiding the inherent defects of optical measurement being affected by light, dust, and vibration. On the other hand, Beidou reference stations are deployed at the base of the piers or towers to dynamically monitor the slow rotation and translation of the erected beam segment caused by subsequent loads, creep, and temperature. The reference conversion algorithm is used to correct the drift of the measurement reference in real time, so that the updated beam segment rotation data is always referenced to the absolute spatial coordinates.

[0035] Simultaneously, the updated pose is compared with the pre-deformation results on a degree-of-freedom basis. The adjustment command is triggered only when the deviation exceeds the threshold; otherwise, it is directly locked, avoiding invalid adjustment actions. This eliminates the long-term cumulative error of the deformation of the erected beam segment on the assembly reference from the source, ensuring that the accuracy of pose perception during cantilever erection does not decrease with the progress of construction. The use of a non-visual sensor array ensures continuous operation under harsh conditions such as high humidity, nighttime, or rain and fog, while the threshold-based instant triggering mechanism significantly reduces the positioning decision delay, providing a highly reliable, real-time, and high-precision pose feedback closed loop for rapid matching and assembly.

[0036] In a preferred embodiment of the present invention, during the segmental assembly process, each execution of the adjustment command causes relative sliding or micro-movement between the bottom surface of the beam segment and the temporary support. This sliding history alters the frictional characteristics, contact stiffness, and gap distribution of the contact interface, resulting in a path-dependent evolution of the frictional state. To identify these time-varying parameters online, S204 is executed at the start of each adjustment action. S204: Receives the adjustment command, analyzes the wear and compaction trend of the support pad contact surface based on the lifting force and displacement response recorded by each outrigger multiple times, identifies the friction state parameters of the current support pad contact surface, and inputs them into the hydraulic control system as feedforward compensation parameters for this adjustment command, so that the system can accurately overcome friction during formal adjustment and avoid displacement lag or overshoot. Specifically, at the start of each adjustment operation, the lifting force and displacement curves of each outrigger are recorded. While the outrigger lifting force gradually increases from zero, but before macroscopic sliding of the beam segment occurs, the thrust sensor continuously records the real-time lifting force. When the lifting force increases to a preset threshold, macroscopic sliding of the beam segment begins. The recorded peak lifting force at this point is the maximum thrust required to overcome static friction. Dividing this peak lifting force by the normal support reaction force at the outrigger, measured by a load cell, yields the static friction coefficient, which reflects the frictional characteristics of the contact interface as it transitions from a static state to the critical sliding state. The normal support reaction force is the normal pressure generated by the beam segment's own weight at the outrigger, measured in real-time by a load cell. Once macroscopic sliding begins in the beam segment, the lifting force will decrease from its peak and tend towards a relatively stable fluctuation range, which can be determined using the envelope method. During the continuous sliding phase, the arithmetic mean of the lifting forces measured by the thrust sensor at various moments is taken to obtain the average lifting force during the sliding phase. Dividing this average lifting force by the normal reaction force at the outrigger gives the dynamic friction coefficient. This coefficient reflects the frictional characteristics of the contact interface under continuous relative motion.

[0037] In the static friction stage before macroscopic sliding occurs in the beam segment, the lifting force and displacement have an approximately linear relationship. Record the difference in lifting force and displacement at the beginning and end points of this stage. Divide the difference in lifting force by the difference in displacement, and the resulting slope is the equivalent contact stiffness of the static friction zone. This parameter reflects the rigidity characteristics of the support material in the elastic compression stage.

[0038] The static friction coefficient obtained from the actual position adjustment is weighted and averaged with the predicted value for the previous position adjustment to obtain the predicted static friction coefficient for the next position adjustment. The weights can be obtained using an online optimization method based on gradient descent. Friction state parameters include static friction coefficient, dynamic friction coefficient, and contact stiffness. Based on the static friction coefficient, a predicted value is obtained and used as a feedforward compensation parameter to generate the next adjustment command, thus offsetting the influence of friction nonlinearity on the adjustment. If the predicted static friction coefficient is significantly higher than the historical average, it indicates that the interface has become rough, requiring a larger pre-thrust to overcome static friction. If the predicted dynamic friction coefficient decreases, it indicates that the interface may have entered a lubricated state, requiring a reduction in slip compensation to avoid overshoot. This predictive mechanism allows the control system to anticipate changes in friction state, rather than responding passively.

[0039] The support contact surface refers to the contact interface between the support leg and the bottom of the beam, and its condition affects the accuracy of displacement response.

[0040] The wear compaction trend refers to the gradual evolution of the surface morphology and mechanical properties of the support contact surface during repeated repositioning processes, caused by the relative sliding or micro-movement between the bottom surface of the beam segment and the support. Specifically, this includes the increase in surface roughness or material spalling due to frictional wear, leading to an upward trend in the static friction coefficient. This trend reflects the path dependence of friction state parameters. As the number of repositioning cycles increases, the static friction resistance, dynamic friction resistance, and contact stiffness are not constant but undergo systematic changes in a certain direction. This method identifies this evolution direction and rate by analyzing historical lifting force and displacement curves, and predicts the friction state during the next repositioning accordingly. This allows for advance adjustment of feedforward compensation parameters, avoiding dead zone crossing failure or slip overshooting loss of control due to neglecting the wear compaction trend.

[0041] S205: Before the formal adjustment, small trial displacements are applied to each leg in sequence. Based on the displacement response deviation of each leg, the displacement coupling relationship between the legs caused by the flexible deformation of the beam segment is identified. During the assembly process, the initial mutual influence parameter matrix for compensating for the flexible influence is established and updated.

[0042] Before the formal positioning, the edge server sequentially sends a small, tentative displacement command to each outrigger, for example, requesting a 0.3 mm lift. The outrigger's controller drives the outrigger to execute the command, while displacement sensors record the actual displacement value of the outrigger to obtain the displacement response deviation. This deviation is used to infer the flexible deformation characteristics of the beam. Specifically, it includes the difference between the commanded displacement and the actual displacement of the active outrigger, reflecting the absorption or amplification of the lift by the beam's flexibility at that outrigger; and the additional displacement generated at other inactive outriggers, reflecting the degree to which the beam's flexible deformation transmits the displacement of the active outrigger to other outriggers. Together, they describe the overall flexible response characteristics of the beam under a unit lift.

[0043] Since the steel-concrete composite beam is a flexible body, when multiple outriggers are lifted synchronously, the displacements and forces between the outriggers are coupled through the bending stiffness of the beam. This coupling effect cannot be described by rigid body kinematics. Therefore, before the initial assembly of the beam segment, the overall deformation response of the beam under a unit outrigger displacement is calculated using offline finite element analysis to establish an initial mutual influence parameter matrix. The elements of this matrix represent the additional displacements generated at other outriggers when a unit displacement is applied to each outrigger. The specific construction process is as follows: a fine model of the beam segment is established in the finite element software. A unit displacement, such as 1 mm, is applied sequentially at each outrigger position. The deformation response of the entire beam is calculated, and the additional displacement values ​​at each outrigger position are extracted. These additional displacement values ​​are then filled into the corresponding positions in the matrix to obtain the initial mutual influence parameter matrix. This matrix is ​​used to compensate for the interference of the flexible deformation of the beam on the pose adjustment when subsequent positioning commands are generated. During assembly, after each formal repositioning, the initial mutual influence parameter matrix is ​​corrected online using the deviation between the commanded displacement and the actual displacement recorded in this repositioning. The correction formula is: New matrix = Old matrix + Learning rate × (Actual displacement deviation vector - Old matrix × Leg reaction force change vector) × Transpose of leg reaction force change vector ÷ (Leg reaction force change vector ∙ Transpose of leg reaction force change vector). As the number of assembly segments increases, the matrix gradually converges to the true value reflecting the actual stiffness distribution of the current beam. This correction formula uses the displacement deviation information observed in this repositioning to infer the error of the matrix and corrects the matrix in the direction of decreasing error.

[0044] The numerator is: (actual displacement deviation vector - old matrix × leg reaction force change vector) × the transpose of the leg reaction force change vector. This represents a matrix obtained by multiplying the residual vector of the predicted displacement deviation using the current matrix by the transpose of the reaction force change vector. This matrix indicates the correction direction of the initial mutual influence parameter matrix. It is the gradient direction in the steepest descent method, enabling the corrected matrix to better fit the observed data. (actual displacement deviation vector - old matrix × leg reaction force change vector) is the prediction residual. The denominator ((outrigger reaction force change vector ∙ outrigger reaction force change vector transpose) represents the inner product of the outrigger reaction force change vectors, i.e., the sum of squares of each component. It is a scalar, numerically equal to the sum of the squares of all outrigger reaction force changes, and physically reflects the total energy or amplitude sum of the squares of all outrigger reaction force changes during this adjustment. This denominator is used to normalize the correction amount, decoupling the correction step size from the intensity of the reaction force changes and ensuring the stability of the algorithm.

[0045] The actual displacement deviation vector is a vector formed by the difference between the commanded displacement value and the actual displacement value of each leg after the formal adjustment is completed. If the difference between the commanded displacement value and the actual displacement value of the leg is positive, it means that the actual displacement is less than the commanded value, reflecting that the beam has flexibly absorbed part of the stroke. If it is negative, it means that there may be slip overshoot or coupling push of other legs. If it is zero, the displacement response is consistent with the commanded value. The outrigger reaction force change vector is a vector composed of the difference between the normal support reaction force of each outrigger after the repositioning is completed and before the repositioning begins. It is used to reflect the redistribution of the load inside the beam during the repositioning process. If the difference between the normal support reaction force after the repositioning is completed and before the repositioning begins is positive, it indicates that the support has borne a larger load after the repositioning, reflecting that the support leg has been lifted or the center of gravity of the beam has shifted to the support leg. If it is negative, it reflects that the support leg has been lowered or the center of gravity of the beam has moved away from the support leg. The learning rate is used to control the correction magnitude of the compliance matrix after each adjustment. The value is usually between 0.05 and 0.2, and the specific value is obtained according to the on-site debugging and optimization method. The larger the value, the greater the impact of the deviation of this adjustment on the matrix correction, and the faster the matrix convergence speed, but it may fluctuate due to the noise of a single measurement. This invention aims to upgrade a hydraulic positioning system from open-loop rigid body motion control to closed-loop mechanical behavior self-learning control. On the one hand, it utilizes the measured lifting force and displacement curves during each positioning to extract the evolution law of static friction coefficient, dynamic friction coefficient and contact stiffness online, and uses an exponentially weighted moving average to predict the friction state of the next positioning, thereby providing feedforward compensation for dead zone crossing and slip overshoot, and further eliminating path dependence deviations caused by support wear, compaction or lubrication during repeated positioning. On the other hand, by actively stimulating the flexible response of the beam through small trial displacements, the displacement coupling matrix between the legs is identified, and the normalized gradient descent method is used to make the matrix converge to the true stiffness distribution as the construction progresses.

[0046] Therefore, its beneficial effects are as follows: friction feedforward suppresses the nonlinearity of the response between the adjustment command and the actual displacement, eliminates dead zone and jump phenomena, and greatly improves the success rate of single adjustment; secondly, the adaptive update of the compliance matrix avoids the coupling mismatch caused by the rigid body assumption, and significantly enhances the coordination accuracy of each leg; then, the combined effect of the two makes the control strategy always close to the real physical state of the beam and pad system. Even after multiple adjustments or replacements of the pads, the system can still maintain relatively consistent dynamic response characteristics, thereby compressing the number of adjustment iterations for cantilever erection to the theoretical lower limit, while eliminating the erosion of the final closure accuracy by the implicit cumulative deviation.

[0047] In a preferred embodiment of the present invention, S301: Based on the comparison results during the cantilever erection process, the theoretical stroke of each leg in collaborative assembly is obtained by combining inverse kinematics solution with dynamic compensation of the initial mutual influence parameter matrix, thereby achieving precise pose matching under the flexible deformation of the beam. Specifically, based on the instantaneous linear mapping relationship between the outrigger geometry and the small displacement of the beam in Cartesian space, the Jacobian matrix is ​​derived analytically, and the inverse of the Jacobian matrix is ​​calculated. The inverse matrix is ​​used to map the six-dimensional pose deviation space to the multi-dimensional outrigger travel space. Then, the inverse matrix is ​​multiplied with the comparison result to obtain the initial travel that each outrigger should produce under the rigid body assumption. Since the steel-concrete composite beam segment is actually a flexible body, the initial travel calculated based on the rigid body assumption will deviate during execution. After one outrigger is lifted, the bending deformation of the beam will cause additional displacement at other outriggers, resulting in the actual pose change deviating from the expectation. Therefore, the system needs to perform flexible compensation correction on the initial travel. The specific correction process is as follows: First, the real-time updated initial mutual influence parameter matrix is ​​called. Then, the system multiplies the current comparison result with the proportional gain matrix to obtain a set of adjustment amounts. Next, the initial mutual influence parameter matrix is ​​multiplied with this set of adjustment amounts to calculate the estimated additional displacement of each outrigger due to the flexible deformation of the beam. Finally, these estimated additional displacements are subtracted from the initial travel to obtain the corrected travel, i.e., the theoretical travel. This corrected travel eliminates the flexible coupling error between outriggers, so that the actual displacement of each outrigger, after being superimposed with the flexible deformation of the beam, can precisely coordinate to produce the target pose change. This correction process is repeated in each positioning control cycle to adapt to the changes in the initial mutual influence parameter matrix as the assembly progresses online.

[0048] The adjustment amount reflects the degree of deviation that the system expects to eliminate through flexible compensation under the current pose deviation; the proportional gain matrix is ​​a diagonal matrix obtained by engineering experience calibration or offline simulation optimization. Each element on its diagonal corresponds to the gain coefficient of the six degrees of freedom of the beam segment. It is used to convert the comparison results into intermediate adjustment amounts for flexibility correction, control the strength of flexible compensation, and the value ranges from 0.1 to 1.0. S302: Based on the feedforward compensation parameters, by calculating the pre-lift force required to overcome static friction and the slip overshoot during the sliding process, dead zone crossing compensation and slip feedforward compensation are performed on the theoretical stroke to obtain the command stroke that eliminates frictional nonlinearity during cantilever erection. The core objective of this step is to eliminate the frictional nonlinearity of the contact interface through feedforward compensation, so that the actual displacement of the outriggers can linearly respond to the command displacement, avoiding dead zone and jumping phenomena. The following is the complete detailed execution analysis process: In dead zone crossing compensation, due to the existence of static friction resistance, when the jacking force applied by the outrigger is less than the static friction resistance, the beam segment will not produce macroscopic displacement, forming a dead zone. Therefore, before the formal displacement control begins, sufficient jacking force needs to be established in advance to cross the dead zone. First, the pre-jacking force is obtained by multiplying the predicted static friction coefficient by the normal reaction force by the safety factor. The purpose of setting this safety factor is to make the thrust infinitely close but not trigger slippage, avoiding unexpected displacement during the dead zone crossing process. It is usually taken as 0.95. Then, the outrigger controller switches to force control mode, using the pre-lifting force as the control input, and drives the servo valve to supply oil to the cylinder. The controller monitors the actual lifting force fed back by the pressure sensor in real time, and adjusts it to gradually approach the pre-lifting force; when the actual lifting force reaches the pre-lifting force, it maintains this lifting force state for 0.5 to 1 second. During this period, the support material undergoes elastic compression, the contact surface gap is eliminated, and the beam segment is in a critical state of about to slide but not yet sliding. After completing the dead zone crossing, the controller switches back to displacement control mode, ready to execute the formal displacement command. At this time, even a small displacement command will be directly converted into the sliding of the beam segment, eliminating the response hysteresis and nonlinearity caused by the dead zone, and reducing impact; Pre-lift force is used to cross the dead zone. Crossing the dead zone refers to the process by which the system overcomes static friction resistance by increasing the lift force, so that the beam segment changes from a static state to a sliding state. The purpose is to eliminate the nonlinear region where there is a command but no response. When a beam segment begins to slide against static friction, the friction state at the contact interface changes from static friction to dynamic friction. In slip feedforward compensation, because dynamic friction resistance is usually less than static friction resistance, the beam segment will experience an acceleration at the moment of sliding, causing the actual displacement to exceed the commanded displacement, resulting in a jump phenomenon and displacement overshoot. Furthermore, fluctuations in friction resistance throughout the sliding process also cause displacement tracking errors. The purpose of slip feedforward compensation is to pre-increase the commanded displacement so that the actual displacement is exactly equal to the target displacement.

[0049] First, based on the theoretical travel and the predicted dynamic friction coefficient, the additional displacement caused by frictional lag during the sliding process is calculated, i.e., the slip overshoot at the moment of sliding. This reflects the severity of the frictional abrupt change and the influence of the beam segment's sliding inertia. The larger the difference between static friction resistance and sliding friction resistance, the more severe the drop in resistance from rest to sliding, and the larger the overshoot. The calculation formula is: Slip overshoot = empirical coefficient × (predicted static friction resistance - predicted sliding friction resistance) ÷ predicted sliding friction resistance × theoretical travel; where the predicted static friction resistance is the predicted value of static friction coefficient × normal reaction force; (predicted static friction resistance - predicted sliding friction resistance) ÷ predicted sliding friction resistance is used to reflect the relative excess of static friction resistance compared to sliding friction resistance. The larger the ratio, the greater the initial acceleration obtained by the beam segment, and the larger the slip overshoot.

[0050] The empirical coefficient is a scaling factor in the calculation of slip overshoot. It is used to convert the degree of frictional abrupt change into displacement overshoot. It can be obtained through engineering experience or offline simulation optimization. For example, empirical values ​​can be directly taken according to the type of support material. The recommended empirical coefficient range for rubber supports is 0.4~0.6, and the recommended empirical coefficient range for steel plates is 0.1~0.3. The slip overshoot is added to the theoretical travel to obtain the commanded travel. The purpose is to make the actual displacement after the slip jump exactly equal to the theoretical travel.

[0051] Static friction resistance refers to the minimum thrust required to overcome the static state of the contact surface, i.e., the minimum lifting force. It determines the critical condition for displacement initiation and is used to set the pre-lift force to cross the dead zone. The predicted sliding friction resistance refers to the continuous lifting force required to maintain relative motion. It affects the smoothness of displacement motion and is used to correct the lifting force to avoid overshoot. The calculation method of its internal sliding friction coefficient prediction value is the same as that of the static friction coefficient prediction value. S303: The commanded stroke of each outrigger during the cantilever erection process is taken as the target displacement. Combined with the allowable bending moment limit of the steel-concrete composite beam segment to be erected in the assembled state, the movement speed of each outrigger during the erection and alignment process at the end face of the beam segment is controlled to obtain the movement trajectory of the position of each outrigger changing with time. S304: Based on the speed of movement, drive the coordinated movement of each outrigger to perform precise alignment between the end face of the cantilever beam segment and the end face of the already erected beam segment.

[0052] To ensure that the beam segment does not generate excessive local stress and dynamic impact during the repositioning process, an S-shaped acceleration / deceleration curve is adopted: the acceleration increases linearly from zero to its maximum value, then decreases symmetrically to zero, with the velocity curve symmetrical about the midpoint of time. The maximum acceleration is calculated based on the support bearing capacity and the allowable bending moment of the beam: the smaller of twice the allowable bending moment ÷ (the square of the outrigger spacing × linear density product) and the difference between the maximum lifting force of the outrigger and the current normal reaction force divided by the effective mass of the beam segment. This step ensures that the dynamic bending moment at the beam section does not exceed the allowable value during the acceleration / deceleration of the outriggers. Using this acceleration as a constraint, an acceleration curve is planned that starts from zero, increases linearly to the peak value, and then decreases linearly back to zero. Integrating the acceleration curve yields the velocity curve, which fully reflects the change law of motion velocity over time. This facilitates precise control of the motion velocity of each outrigger at each moment, thereby ensuring smooth alignment of the beam segment without impact. One aspect is based on Newton's second law, starting from the outrigger's own output capacity, calculating the maximum acceleration that the outrigger can generate, and obtaining the remaining available lifting force margin of the outrigger by the difference between the maximum lifting force of the outrigger and the current normal reaction force; the maximum lifting of the outrigger is the upper limit of the output force designed by the hydraulic system, which can be directly read from the equipment nameplate; The effective mass reflects the share of inertial load borne by the outrigger, and its value is equal to the normal reaction force divided by the gravitational acceleration. On the other hand, based on the relationship between bending moment and acceleration of simply supported beams in structural mechanics, starting from the load-bearing capacity of the beam structure, the maximum allowable acceleration is calculated under the premise that the beam does not produce plastic deformation. The numerator is multiplied by 2 because in the simply supported beam model, the maximum bending moment of the beam under uniformly distributed acceleration load is proportional to the acceleration, and a coefficient needs to be multiplied when solving the inverse problem. The denominator reflects the product of the square of the leg spacing and the linear density of the beam segment, which describes the geometric and mass distribution characteristics of the inertial bending moment generated by the beam segment under unit acceleration. The larger the value, the larger the bending moment generated by the same acceleration, and therefore the smaller the allowable displacement acceleration.

[0053] The square of the outrigger spacing reflects the square effect of the lever arm. The larger the outrigger spacing, the greater the bending moment generated by the same acceleration, and therefore the smaller the allowable acceleration. The outrigger spacing is the horizontal distance between two adjacent outriggers. Linear density is the mass per unit length of a beam segment, reflecting the density of mass distribution along the beam length. The unit is kg / m, which can be directly read from the design drawings. The allowable bending moment is the upper limit of the cross-sectional bending capacity set during the cantilever erection process to ensure that the steel-concrete composite beam segment does not undergo plastic deformation or structural damage due to excessive local bending moment. It can be directly read from the design drawings. In this embodiment of the invention, the rigid body theoretical stroke is modified into an elastic compensation stroke that adapts to the flexible deformation of the beam by using the Jacobi inverse matrix and the real-time updated flexibility matrix, thereby eliminating displacement coupling distortion between the legs; then, based on the online identified static and dynamic friction resistance, feedforward compensation is performed on dead zone crossing and slip overshoot respectively, so that the actual displacement response is close to linear, and the nonlinear jump caused by support wear or lubrication is suppressed. Finally, using the allowable bending moment of the beam as a hard constraint and combining it with the force output capacity of the outriggers, an S-shaped velocity trajectory was planned to ensure that the dynamic bending moment of each section does not exceed the limit during acceleration and deceleration.

[0054] Therefore, the positioning command simultaneously possesses mechanical rationality, smooth motion, and structural safety. It eliminates the erosion of positional accuracy by flexible coupling and frictional nonlinearity, and transforms dynamic bending moment constraints from passive verification to active planning, enabling cantilever erection to leap from trial-and-error approximation to precise alignment.

[0055] In a preferred embodiment of the present invention, S401: During the cantilever erection, additional bending and shear stresses will be generated inside the beam. To prevent local stresses from exceeding the tensile strength of concrete or the yield strength of steel beams, the normal reaction force of each leg and the strain data of the current steel-concrete composite beam section are monitored in real time. By analyzing the stress balance of the legs and the structural stress safety state, the leg displacement fine-tuning amount is generated and stress coordination processing is performed to ensure that the stress distribution of the beam is uniform and does not exceed the material safety limit during the entire repositioning process. S4011: The positioning action is performed based on the target displacement of each outrigger; S4012: Calculate the average reaction force of each outrigger based on the normal reaction force of each outrigger, compare the difference between the normal reaction force of a single outrigger and the average reaction force, and obtain the displacement fine adjustment amount of the corresponding outrigger by combining the preset equalization gain coefficient. Add the displacement fine adjustment amount to the original adjustment action to achieve balanced force adjustment between outriggers during cantilever erection. Based on the load distribution principle of multi-point supported continuous beams in structural mechanics, this method calculates the arithmetic mean of the normal reactions of all outriggers. The current normal reaction is then compared to the average reaction to obtain the deviation between the current outrigger's reaction and the average. This deviation is multiplied by a preset equilibrium gain coefficient to obtain the displacement fine-tuning amount for that outrigger. Each outrigger, according to the sign and magnitude of its displacement fine-tuning deviation, adds a corresponding fine-tuning amount to its original target displacement: outriggers with reactions greater than the average receive a negative fine-tuning amount, indicating a slight decrease; outriggers with reactions less than the average receive a positive fine-tuning amount, indicating a slight increase. All outriggers simultaneously perform this fine-tuning action, bringing the reactions of each outrigger closer to the average, thereby achieving force equilibrium.

[0056] The displacement fine-tuning is used to make real-time, minute corrections to the displacement commands of each leg while keeping the overall position of the beam segment basically unchanged. This reduces the difference in reaction force between the legs, making the stress distribution inside the beam more uniform. It also prevents local cracking or yielding of the beam due to overload of a certain leg and avoids additional bending deformation of the beam due to uneven stress, thereby ensuring the structural safety of the adjustment process and the final assembly accuracy.

[0057] The equalization gain coefficient is the proportional gain parameter of the stress equalization controller. The smaller the coefficient, the smaller the adjustment force and the slower the convergence. It is used to convert the outrigger reaction force deviation into displacement fine adjustment, control the response intensity of stress equalization, and avoid oscillation caused by over-adjustment. Its value is obtained through engineering experience, on-site debugging or offline simulation, and is usually between 0.3 and 1.2. Its unit is mm / kN. The ultimate goal of this method is to minimize the variance of the reaction forces at each support point, so that all legs are subjected to forces as uniformly as possible, thereby preventing local stresses from exceeding the tensile strength of concrete or the yield strength of steel beams.

[0058] S4013: Preset outage reaction force thresholds and beam strain thresholds. The real-time monitored normal reaction force and strain data are compared with the corresponding thresholds to determine if the structural stress is within a safe range. If any monitored data exceeds the corresponding threshold, the assembly and repositioning process is immediately paused, and all outriggers are simultaneously subjected to minor unloading to allow the stress in the current steel-concrete composite beam segment to redistribute evenly, completing the unloading of this cantilever erection segment. For example, if the normal reaction force exceeds the outage reaction force threshold, the assembly and repositioning process is paused. Strain data are obtained using resistance strain gauges; When any monitored data exceeds the safety threshold, a synchronous unloading command is immediately generated. This command includes the unloading displacement, such as a descent of 0.5 mm, the movement speed, such as 0.2 mm / s, and a global synchronization timestamp. Upon receiving the command, each controller calculates the waiting time based on the unified timestamp to ensure all outriggers begin unloading simultaneously. Upon reaching the synchronization time, each outrigger descends at the commanded speed at a uniform rate, with displacement sensors providing real-time position feedback to ensure synchronization accuracy. Once all outriggers reach the target, it is confirmed that the reaction force of each outrigger is below the corresponding threshold, and unloading is complete. If there is still an overshoot, a second round of micro-unloading is performed, with the cumulative descent not exceeding a preset maximum value, typically 2 mm. S402: After completing this assembly, unloading or cantilever erection, re-execute S201 to S203 to determine whether the adjustment command is triggered. If the deviation meets the fine assembly requirements, terminate the matching assembly operation of the current steel-concrete composite beam segment and perform the contact state reset procedure on the support leg corresponding to the current steel-concrete composite beam segment, that is, enter the friction state reset to eliminate the residual friction stress of the adjustment. This step applies to situations where the pose needs to be remeasured and the adjustment needs to continue, including situations where the threshold is not met after normal adjustment is completed, and situations where the adjustment process is restored after unloading due to exceeding the limit.

[0059] The contact state reset process includes synchronously lifting all outriggers to detach the beam segment from the support pad, maintaining the lifting position while applying high-frequency micro-amplitude vibration to eliminate plastic indentations and debris adhesion at the contact interface, and finally synchronously lowering the beam segment to re-engage with the support pad, thereby restoring the initial frictional characteristics of the contact interface. For example, each outrigger is required to be lifted upwards by 1.0 mm at a slow speed of 0.2 mm / s. Once all outriggers have reached the target lifting height, this lifting state is maintained for 5 seconds. During this period, the controller of each outrigger performs vertical reciprocating vibration at a high frequency of 2 Hz and a small amplitude of ±0.1 mm. After the vibration ends, each outrigger is required to descend back to its original support height at the same slow speed of 0.2 mm / s as the lifting. In this invention, all parameters are dimensionless by using dimensionless processing technology to remove their dimensions, and all thresholds can be obtained by the mean-standard deviation method. S403: If the deviation does not meet the fine assembly requirements and the number of adjustments has not exceeded the adjustment threshold, the adjustment command will be regenerated, and the next round of adjustment will be performed on the current steel-concrete composite beam segment until the deviation meets the fine assembly requirements and the adjustment stops, and the reset is completed. If the number of adjustments exceeds the limit, the action will stop and an alarm will be triggered, indicating that the current structural state cannot be corrected by conventional iteration. All outrigger actions will be stopped immediately and an audible and visual alarm will be triggered to prompt the on-site operator to intervene and check. After the operator has checked for problems such as structural jamming, abnormal settlement of support points, deviation of measurement benchmarks, or equipment failure, he / she can choose to reset and restart the adjustment process or switch to manual intervention fine adjustment mode. Automatic assembly operation will continue after the abnormality is eliminated in order to protect the beam from damage caused by repeated hard adjustments.

[0060] This invention aims to unify the structural safety and precision of cantilever erection within the same closed-loop frame. Based on the load distribution principle of multi-point supported continuous beams, it monitors the normal reaction force of each leg in real time and calculates its deviation from the average value. The reaction force deviation is converted into a displacement fine-tuning amount using an equalization gain coefficient, so that each leg can be slightly raised or lowered while maintaining the overall posture, thereby minimizing the reaction force variance and avoiding local stress concentration from the source. At the same time, a dual safety threshold of reaction force and strain is preset. Once the limit is exceeded, all outriggers are immediately triggered to unload in a synchronous, equal speed and equal amount. A global timestamp is used to ensure synchronization, so that the stress of the beam is quickly redistributed without generating a secondary impact. This is equivalent to equipping the repositioning process with a real-time stress melting mechanism. After each repositioning or unloading, the system re-measures the pose and enters the branch decision. If it is qualified, the contact state reset procedure is executed to eliminate the plastic indentation and friction path dependence accumulated by repeated repositioning. If it is unqualified but not exceeding the limit, it automatically iterates and readjusts. If it exceeds the limit, it alarms and stops to avoid blindly adjusting and damaging the beam.

[0061] In a preferred embodiment of the present invention, S501: After the readjustment and assembly are completed, the gap values ​​at the eddy current displacement sensors in the four quadrants of the end face of the current steel-concrete composite beam segment are measured. The independent preload required for each bolt or weld point is calculated through the elastic contact mechanics relationship. The target preload for each bolt or weld point is controlled to obtain the preload command set. In practice, once the repositioning and retesting after resetting meet the precision assembly requirements, the system prepares for mechanical locking. Due to potential residual errors in the prediction of end-face humidity shrinkage distortion, the actual end-face contact pressure distribution may be uneven.

[0062] Therefore, before locking, the gap between the end face of the beam segment to be assembled and the end face of the already erected beam segment is measured using eddy current displacement sensors in the four quadrants of the end face. The average gap value is calculated as the reference gap. It is assumed that there are a total of several connecting bolts, evenly distributed around the perimeter of the end face. According to Hertzian contact theory in elastic contact mechanics and the basic principle of bolt connection, the theoretical preload at the corresponding bolt position should be inversely proportional to the local gap at that location. Specifically, the gap value corresponding to the bolt position is read, a very small positive number is added to this gap value to prevent the denominator from being zero, and then the reference gap is divided by this result to obtain a ratio. This ratio is multiplied by the reference preload specified in the design drawings to obtain the target preload at that bolt position. This target preload is inversely proportional to the local gap, that is, the smaller the gap, the greater the preload; the larger the gap, the smaller the preload.

[0063] Subsequently, the target preload of each bolt is multiplied by a torque coefficient, and then multiplied by the nominal diameter of the bolt to obtain the target torque value for that bolt. After summing the target torque values ​​of all bolts, a preload command set is generated and sent to automated connection equipment, such as a hydraulic wrench or an electric torque wrench. The torque coefficient reflects the efficiency of torque conversion into preload during bolt tightening; it is measured using a laboratory torque coefficient measuring instrument or provided by the supplier. The target preload reflects the axial tensile force required to ensure uniform contact of the end faces at each bolt position when the end face gap is unevenly distributed. The smaller the gap, the greater the preload required to overcome the local protrusion. The target torque value is the tightening torque applied to the bolt, which serves as the control command for a hydraulic or electric torque wrench. S502: Based on the pre-tightening instruction set, the connecting bolts at the end face of the current steel-concrete composite beam segment are tightened or welded in sequence. During the tightening process, the system monitors the deviation between the actual torque and the target torque of each bolt in real time. If the deviation exceeds ±3%, it will automatically tighten the bolts. Before welding, the system applies reverse pre-deformation through the support legs to complete the deformation compensation and locking during the welding connection process. For welded joints, the expected shrinkage deformation caused by welding heat input is calculated. First, welding tests are conducted on test plates using the same steel plate thickness, welding process parameters, and joint type as the actual beam segment. Before welding the test plates, displacement sensors are placed symmetrically on both sides of the weld to measure the deformation changes during welding and after cooling. The average value of multiple tests is taken to obtain the shrinkage deformation per unit length of the weld under these process conditions. Therefore, for a weld of length L, the expected shrinkage deformation is the product of the weld length and the shrinkage deformation per unit length of the weld. The expected shrinkage deformation is then reversed to obtain the reverse pre-deformation required before welding. Reverse pre-deformation is an active attitude correction applied before welding. It is set according to the predicted amount of cooling shrinkage deformation, with the aim of canceling out the deformation after cooling. S503: After the current steel-concrete composite beam segment is quickly assembled and cooled, the beam segment rotation data of the current steel-concrete composite beam segment is re-measured and compared with the pre-deformation results. The residual installation error of the current steel-concrete composite beam segment during the matching assembly process is statistically analyzed to control the assembly benchmark of the next steel-concrete composite beam segment to be erected during the quick matching assembly.

[0064] The pre-deformation results are compared with the beam segment rotation data to obtain the residual installation error, which includes the deviation values ​​in six directions: front-back translation, left-right translation, up-down lifting, pitch, yaw, and roll. This reflects the gap between the actual installation position of this segment and the theoretical target after all the adjustment, welding, and locking processes. Based on the residual installation error, the assembly datum for the next steel-concrete composite beam segment to be erected during rapid matching and assembly is controlled. Specifically, the assembly datum for the next segment = theoretical design datum + initial deviation compensation of the previous segment + compensation coefficient × residual installation error of the current segment. This calculation process weights and adds the compensation experience accumulated from the previous segment to the newly generated residual error to form the initial deviation compensation value for the assembly of the next segment.

[0065] The theoretical design benchmark is a static, permanent ideal alignment, which can be obtained directly by reading the segment installation coordinate table in the bridge construction design drawings or the BIM model; The initial deviation compensation amount of the previous segment + compensation coefficient × residual installation error of the current segment represents the initial deviation compensation amount when assembling the next segment. This facilitates the gradual convergence of the assembly target of subsequent segments towards the actual assembled alignment, realizing the step-by-step transmission and smooth correction of errors, and avoiding excessive deviations in the closure segment.

[0066] The compensation coefficient is less than 1, and is usually 0.8, to avoid overcompensation. The specific value can be determined based on engineering experience or offline simulation optimization.

[0067] The assembly benchmark reflects the target position and orientation that the steel-concrete composite beam segment to be erected should achieve in space. It is a control target determined comprehensively based on factors such as theoretical design alignment, cumulative error of assembled segments, and prediction of end-face humidity distortion. In this embodiment of the invention, based on the measured gaps in the four quadrants of the end face, the target preload of each bolt is inversely proportional to the local gap according to the Hertzian contact theory. The smaller the gap, the greater the preload. This transforms the uneven contact caused by humidity distortion into uniform surface pressure, avoiding the hidden defects of overpressure on protrusions and suspension of concave areas under traditional uniform preload. Next, the shrinkage per unit length of the welded connection was calibrated by process test, and reverse pre-deformation was applied in advance so that the shrinkage of the weld after cooling was exactly offset, and the position was locked so that it would not drift due to heat input. Finally, after the segment cools down, the final pose is remeasured, the residual installation error is calculated, and the assembly reference of the next segment is updated in a recursive manner, so that the target pose of the subsequent segments gradually approaches the actual assembled line shape.

[0068] Thus, this invention achieves a leap from single-point locking to full-bridge alignment self-adaptation. Non-uniform pre-tightening eliminates local stress concentration and long-term fatigue risks at the end face; reverse welding pre-deformation transforms thermal effects from a source of disturbance into a negligible amount; and error recursive compensation transmits and attenuates the residual deviation of each segment in a controllable proportion, ensuring that deviations in the closure segment do not accumulate and diverge. The synergy of these three elements enables the cantilever erection of large-span steel-concrete composite beams to possess both millimeter-level single-segment locking accuracy and full-bridge-scale alignment self-healing capability.

[0069] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for rapid matching and assembly control of segments in a large-span steel-concrete composite beam, characterized in that, The method includes: Obtain humidity data and original beam condition data of the end face of the steel-concrete composite beam segment to be erected, determine the distortion displacement and rotation state of the beam segment end face, and generate pre-deformation results; During the cantilever erection of the bridge, the rotation state of the beam segment is identified and updated. Combined with the pre-deformation results, the triggering of the adjustment command is determined, and the friction state parameters of the contact interface are identified. The displacement coupling relationship between the legs is identified by probing the displacement. Based on the adjustment command and friction state parameters, the position of each outrigger is adjusted to achieve precise alignment of the beam segment end face. The adjustment includes the command stroke and erection speed state of the outrigger coordinated operation. During the erection and alignment process, stress coordination processing is performed based on the force data of the outriggers and the strain data of the beam to complete the unloading of the cantilever erection segment and re-determine the triggering of the adjustment command to form a branch assembly mechanism, including reset locking and readjustment assembly. After the erection and alignment are completed, the connecting bolts at the end face of the current steel-concrete composite beam segment are tightened or welded, and the assembly reference of the next steel-concrete composite beam segment to be erected is recorded.

2. The method for rapid matching and assembly control of large-span steel-concrete composite beam segments according to claim 1, characterized in that, Obtain humidity data and original beam condition data of the end face of the steel-concrete composite beam segment to be erected, determine the distortion displacement and rotation state of the beam segment end face, and generate pre-deformation results, including: Humidity data was obtained by recording the relative humidity changes at various measuring points on the end face of the steel-concrete composite beam segment to be erected; Based on humidity data, the current humidity gradient distribution is calculated in real time, and the humidity diffusion state within the bridge material is analyzed. The uneven shrinkage trend of the end face during the future assembly window before the beam segment is stored and cantilevered is erected is predicted, and the distortion displacement value at each position of the end face is obtained. Based on the distortion displacement value, the corresponding reverse compensation amount for each outrigger is set in the hydraulic control system to identify the pre-deformation result, and the drive end face is aligned with the pre-deformation result during cantilever erection.

3. The method for rapid matching and assembly control of large-span steel-concrete composite beam segments according to claim 2, characterized in that, During the cantilever erection of the bridge, the rotation state of the beam segment is identified and updated. Combined with pre-deformation results, the triggering of the adjustment command is determined. The friction state parameters of the contact interface are identified, and the displacement coupling relationship between the outriggers is determined through trial displacement, including: Based on the original beam state data of the steel-concrete composite beam segment to be erected, the rotation state of the steel-concrete composite beam segment to be erected is analyzed to obtain the beam segment rotation data. Beidou reference stations are set up on the piers or the base of the towers to monitor the rotational displacement of the erected beam segments in real time during the cantilever erection process. The measurement reference is corrected by reference conversion and the beam segment rotation data is updated. During the cantilever erection process, the updated beam segment rotation data is compared with the pre-deformation results. If the comparison result exceeds the preset comparison threshold, it indicates that the deviation does not meet the fine assembly requirements, and the adjustment command is triggered.

4. The method for rapid matching and assembly control of large-span steel-concrete composite beam segments according to claim 3, characterized in that, During the cantilever erection of bridges, the rotation state of the beam segments is identified and updated. Combined with pre-deformation results, the triggering of adjustment commands is determined. Contact interface friction parameters are identified, and displacement coupling relationships between outriggers are determined through trial displacement. Other aspects include: Upon receiving the adjustment command, based on the lifting force and displacement response recorded from multiple adjustments of each outrigger, the wear and compaction trend of the outrigger contact surface is analyzed, and the friction state parameters of the current outrigger contact surface are identified and input into the hydraulic control system as feedforward compensation parameters for this adjustment command. Small trial displacements are applied to each leg in sequence. Based on the displacement response deviation of each leg, the displacement coupling relationship between the legs caused by the flexible deformation of the beam segment is identified. During the assembly process, an initial mutual influence parameter matrix for compensating for the flexible effect is established and updated.

5. The method for rapid matching and assembly control of large-span steel-concrete composite beam segments according to claim 4, characterized in that, Based on the positioning command and friction state parameters, the position of each outrigger is adjusted to achieve precise alignment of the beam segment end face. This adjustment includes the command stroke and erection speed of the outriggers working together, including: Based on the comparison results during the cantilever erection process, the theoretical travel distance of the assembly is obtained by combining inverse kinematics solution with dynamic compensation of the initial mutual influence parameter matrix. Based on the feedforward compensation parameters, by calculating the pre-lift force required to overcome static friction and the slip overshoot during the sliding process, dead zone crossing compensation and slip feedforward compensation are performed on the theoretical stroke to obtain the command stroke that eliminates friction nonlinearity during cantilever erection. The commanded stroke of each outrigger during the cantilever erection process is taken as the target displacement, and combined with the allowable bending moment limit of the steel-concrete composite beam segment to be erected in the assembled state, the movement speed of each outrigger during the erection and alignment process at the end face of the beam segment is controlled. Based on the speed of movement, the outriggers are driven to move in coordination to align the end face of the cantilever beam segment with the end face of the already erected beam segment.

6. The method for rapid matching and assembly control of large-span steel-concrete composite beam segments according to claim 5, characterized in that, During the erection and alignment process, stress coordination processing is performed based on the outrigger stress and beam strain data to complete the unloading of the cantilever erection segment. The triggering of the repositioning command is then reassessed to establish a branch assembly mechanism, including: During this cantilever erection, the normal reaction force of each outrigger and the strain data of the current steel-concrete composite beam section are monitored in real time. By analyzing the stress balance of the outrigger and the structural stress safety state, the outrigger displacement fine adjustment is generated and stress coordination processing is performed. After completing the assembly, unloading or cantilever erection, re-evaluate whether the adjustment command has been triggered. If the deviation meets the fine assembly requirements, terminate the matching assembly operation of the current steel-concrete composite beam segment, and perform a contact state reset procedure on the support leg corresponding to the current steel-concrete composite beam segment to perform a reset and locking operation. If the deviation does not meet the precision assembly requirements and the number of repositioning attempts has not exceeded the repositioning threshold, a repositioning instruction will be regenerated, and the current steel-concrete composite beam segment will be repositioned and reassembled.

7. The method for rapid matching and assembly control of large-span steel-concrete composite beam segments according to claim 6, characterized in that, By analyzing the stress balance of the outriggers and the structural stress safety state, fine-tuning of outrigger displacement is generated and stress coordination processing is performed, including: The repositioning action is performed based on the target displacement of each leg. The average reaction force of each outrigger is calculated based on the normal reaction force of each outrigger. The difference between the normal reaction force of a single outrigger and the average reaction force is compared. The displacement adjustment amount of the corresponding outrigger is obtained by combining the preset equalization gain coefficient. The displacement adjustment amount is superimposed on the original adjustment action to achieve balanced force adjustment between outriggers during cantilever erection. The outrigger reaction force threshold and beam strain threshold are preset. The real-time monitored normal reaction force and strain data are compared with the corresponding thresholds. When any monitored data exceeds the corresponding threshold, the assembly and adjustment action is immediately suspended, and all outriggers are controlled to perform micro-unloading synchronously to complete the unloading of this cantilever erection segment.

8. The method for rapid matching and assembly control of large-span steel-concrete composite beam segments according to claim 7, characterized in that, After the erection and alignment are completed, the connecting bolts at the end face of the current steel-concrete composite beam segment are tightened or welded, and the assembly reference of the next steel-concrete composite beam segment to be erected is recorded, including: After the repositioning and assembly are completed, the end face gap value of the current steel-concrete composite beam segment is measured, the target preload of each bolt or weld point is controlled, and the preload command set is obtained. Based on the pre-tightening instruction set, the connecting bolts at the end face of the current steel-concrete composite beam segment are tightened or welded in sequence, and the reverse pre-deformation is applied by the support leg before welding to complete the deformation compensation and locking during the welding connection process. After the current steel-concrete composite beam segment is quickly assembled and cooled, the beam segment rotation data of the current steel-concrete composite beam segment is re-measured and compared with the pre-deformation results to record the assembly benchmark for the next steel-concrete composite beam segment to be erected during rapid matching assembly.

Images