Large-span steel beam without pier multi-point synchronous step-by-step pushing construction system

Abstract

The application relates to the technical field of construction control, and discloses a large-span steel beam non-pier-supported multi-point synchronous step-by-step incremental launching construction system. The system comprises a measurement module, a configuration module, an assembling module, an establishment module, an incremental launching module and a correction module. The measurement module is used for establishing a control network and generating a construction control database; the configuration module is used for constructing a hierarchical distributed control architecture; the assembling module is used for establishing a steel beam digital model; the establishment module is used for constructing a neural network prediction model; the incremental launching module is used for executing synchronous incremental launching; and the correction module is used for position measurement and correction. Through the establishment of a high-precision control network, the construction of a hierarchical distributed control architecture, the adoption of a neural network prediction model and a feedforward-feedback compound control strategy, high-precision synchronous control of each incremental launching point is realized, the stress of the steel beam is ensured to be uniform during the incremental launching process, steel beam deformation and safety hazards caused by insufficient synchronous precision are avoided, and the safety and accuracy of the non-pier-supported incremental launching construction of the large-span steel beam are significantly improved.

Description

Technical Field

[0001] The present application relates to the field of construction control technology, and in particular to a large-span steel beam pier-free multi-point synchronous walking jacking construction system. Background Art

[0002] Steel girder bridges are a key type of modern bridge construction, particularly long-span steel girder bridges, which offer advantages such as light weight, large spans, and easy construction. They are widely used in transportation projects such as railways, highways, and urban interchanges. Traditional methods for installing steel beams include hoisting, jacking, and rotation. Jacking has become a primary method for constructing long-span steel girder bridges due to its minimal traffic disruption, high safety, and adaptability to a wide range of terrain conditions. Traditional jacking typically involves jacking at a single or limited number of points, with the beams being moved to their designed positions along pre-installed temporary piers and slideways. In situations such as crossing rivers, deep valleys, or existing traffic arteries, the installation of temporary piers often presents challenges such as high construction difficulty, high costs, and significant environmental impact. To address these issues, the engineering community has developed pierless jacking technology, which reduces the need for temporary piers by increasing the number of jacking points, employing a support frame system, or employing cantilever assembly. However, pierless jacking construction places higher demands on the jacking system's load-bearing capacity, synchronization accuracy, and control technology.

[0003] The existing pierless jacking technology still has many shortcomings. First, the synchronization accuracy of the multi-point jacking system is insufficient, and the displacement, speed and thrust between the jacking points are difficult to coordinate accurately, resulting in additional stress and even deformation damage to the steel beam during the jacking process. Secondly, the traditional walking jacking device has poor stability and is prone to offset or uneven settlement under large load conditions, affecting the jacking accuracy. Third, the real-time and reliability of the control system are insufficient, and it is impossible to accurately control and adjust the multi-point jacking process under complex working conditions. Especially when faced with uneven foundations, changes in steel beam stiffness, interference from environmental factors, etc., it is easy to cause safety hazards such as offset and torsion during the jacking process. In addition, the existing technology lacks the ability to accurately monitor and analyze data throughout the jacking process, making it difficult to achieve intelligent control and quality assurance of the jacking process. These technical deficiencies have seriously restricted the application and development of pierless jacking construction of large-span steel beams. Summary of the Invention

[0004] This application provides a multi-point synchronous walking-type jacking construction system for large-span steel beams without piers, which is used to solve the key technical problem of insufficient synchronization accuracy of multi-point jacking in the existing technology. By establishing a high-precision control network, constructing a hierarchical distributed control architecture, and adopting a neural network prediction model and a feedforward-feedback composite control strategy, high-precision synchronization control of each jacking point is achieved, and the synchronization error is controlled within ±0.5mm, ensuring that the steel beams are evenly stressed during the jacking process, avoiding deformation of the steel beams and safety hazards caused by insufficient synchronization accuracy, and significantly improving the safety and accuracy of large-span steel beam jacking construction without piers.

[0005] The present application provides a large-span steel beam without piers and a multi-point synchronous walking-type jacking construction system, which comprises:

[0006] The measurement module is used to measure the construction site, establish a control network and elevation control network, arrange displacement monitoring points, and generate a construction control database;

[0007] A configuration module is used to configure a main control unit and sub-control units based on the construction control database and adopt a hierarchical distributed control architecture to form a real-time monitoring and control platform;

[0008] An assembly module is used to assemble the steel beams according to the numbering sequence under the guidance of the control network, establish a digital model of the steel beams using three-dimensional laser scanning, and determine the geometric parameters of the steel beams;

[0009] Establishing a module for performing parameter optimization on the hierarchical distributed control architecture and establishing a neural network prediction model based on the geometric parameters of the steel beam and the data collected by the real-time monitoring and control platform;

[0010] A pushing module is used to perform multi-point synchronous pushing of steel beams based on the neural network prediction model, adopt a feedforward-feedback composite control strategy, and dynamically adjust the parameters of each pushing point through the real-time monitoring and control platform;

[0011] The correction module is used to measure the spatial position of the steel beam after the jacking is completed using the control network, calculate the position deviation according to the design parameters in the construction control database, and perform position correction according to the position deviation.

[0012] In the technical solution provided by the present application, the large-span steel beam pier-free multi-point synchronous walking jacking construction method of the present invention provides accurate basic data support for subsequent construction by precisely measuring the construction site, establishing a control network and an elevation control network, laying out displacement monitoring points, and generating a construction control database, so that the closure error of the control network is controlled within 1mm / km, and the elevation closure error is controlled within 0.5mm / L, ensuring the high precision of the construction benchmark and effectively avoiding the jacking deviation caused by the accumulation of measurement errors; based on the construction control database, a hierarchical distributed control architecture is adopted, and a main control unit and a sub-control unit are configured to form a real-time monitoring and control platform, achieving a processing capacity of 10GFLOPs. High-performance computing with a clock synchronization accuracy of more than 1μs and precise timing control with a clock synchronization accuracy better than 1μs have greatly improved the system's real-time response capability and control accuracy, keeping communication delays within 10ms and ensuring the timeliness of multi-point coordinated control. Steel beams are assembled in numerical order under the guidance of the control network. 3D laser scanning is used to establish a digital model of the steel beams, determine the geometric parameters of the steel beams, and obtain high-precision 3D data with a point cloud density of 1 point / cm². This achieves accurate quantification of the geometric dimensions of the steel beams and provides a reliable basis for optimizing the jacking strategy. Parameters of the hierarchical distributed control architecture are optimized, and a neural network prediction model is established based on the geometric parameters of the steel beams and data collected by the real-time monitoring and control platform.

[0013] By combining a BP neural network with the Levenberg-Marquardt algorithm, a high-precision system dynamic response prediction model with a prediction error of less than 1% was constructed, significantly improving the control system's adaptability to complex working conditions. Based on the neural network prediction model, a feedforward-feedback composite control strategy was employed to execute multi-point synchronous jacking of steel beams. The parameters of each jacking point were dynamically adjusted via a real-time monitoring and control platform, keeping the synchronization error within ±0.5mm. This ensured uniform force distribution during the jacking process, effectively avoiding beam deformation and structural damage caused by synchronization errors, and significantly improving the safety and accuracy of pier-free jacking of large-span steel beams. The control network was used to measure the spatial position of the steel beams after jacking. The position deviation was calculated based on the design parameters in the construction control database, and position correction was performed based on the position deviation. An adjustment system consisting of precision jacks and wedge-shaped pads was used to achieve high-precision position correction with an adjustment accuracy of 0.1mm, ensuring that the planar position deviation of the steel beams was within ±5mm and the elevation deviation was within ±3mm upon final placement, significantly improving project quality and structural safety. The present invention makes a particularly significant contribution to the application of artificial intelligence algorithms. By combining a neural network prediction model with a feedforward-feedback composite control strategy, it achieves precise modeling and control of complex nonlinear systems during the jacking process of large-span steel beams. The BP neural network can accurately predict the dynamic behavior of the jacking system by learning the response characteristics of the system under different working conditions. The introduction of the Levenberg-Marquardt training algorithm significantly improves the convergence speed and prediction accuracy of the neural network. This intelligent control method with a neural network as its core greatly enhances the ability of the jacking system to cope with complex working conditions such as changes in steel beam weight, uneven foundations, and interference from environmental factors. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0015] Figure 1 This is a schematic diagram of an embodiment of a large-span steel beam pier-free multi-point synchronous walking jacking construction system in the embodiment of the present application. DETAILED DESCRIPTION

[0016] An embodiment of the present application provides a large-span steel beam pier-free multi-point synchronous walking jacking construction system. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or apparatus that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or that are inherent to these processes, methods, products or apparatus.

[0017] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In the embodiment of the present application, an embodiment of a large-span steel beam pier-free multi-point synchronous walking jacking construction system includes:

[0018] The measurement module 101 is used to measure the construction site, establish a control network and an elevation control network, arrange displacement monitoring points, and generate a construction control database;

[0019] Configuration module 102, configured to configure a main control unit and sub-control units based on the construction control database using a hierarchical distributed control architecture to form a real-time monitoring and control platform;

[0020] An assembly module 103 is used to assemble the steel beams according to the numbering sequence under the guidance of the control network, establish a digital model of the steel beams using three-dimensional laser scanning, and determine the geometric parameters of the steel beams;

[0021] Establishing module 104 for optimizing parameters of the hierarchical distributed control architecture and establishing a neural network prediction model based on the geometric parameters of the steel beam and the data collected by the real-time monitoring and control platform;

[0022] A pushing module 105 is configured to execute multi-point synchronous pushing of steel beams based on the neural network prediction model, adopt a feedforward-feedback composite control strategy, and dynamically adjust the parameters of each pushing point through the real-time monitoring and control platform;

[0023] The correction module 106 is used to measure the spatial position of the steel beam after the jacking is completed using the control network, calculate the position deviation according to the design parameters in the construction control database, and perform position correction according to the position deviation.

[0024] Specifically, the measurement module 101 uses a multi-channel geological radar to survey the foundation beneath the jacking track, obtain foundation bearing capacity data, and analyze foundation stability in real time. In practical applications, this module first uses a high-precision total station to establish a construction control network, setting at least eight control points to form a closed measurement network, with a closure error of less than 1 mm / km. A digital level is then used to establish an elevation control network, ensuring an elevation closure error of less than 0.5 mm / L (where L is the distance between the measuring stations, in kilometers). Displacement monitoring points and settlement monitoring points are deployed along the jacking track, with the former spaced 50 meters apart and the latter at 30 meters apart, to collect track deformation data in real time. This data is aggregated and analyzed using specialized software to construct a construction control database containing foundation parameters, control network parameters, and displacement monitoring data. For example, in a large-span steel beam jacking project, geological radar surveys revealed a soil density of 92% beneath the track, a bearing capacity of 320 kPa, and an elevation control network closure error of 0.3 mm / km, meeting construction accuracy requirements.

[0025] The configuration module 102 builds a hierarchical distributed control architecture based on the construction control database. The architecture includes a master control unit and multiple sub-control units. The master control unit uses an industrial computer with a processing capacity of more than 10 GFLOPS, and sets up a redundant backup system to prevent data loss. The sub-control units are composed of programmable logic controllers, with one sub-control unit configured at each jacking point to execute instructions from the master control unit and feed back real-time data. The master control unit and the sub-control units are connected through an industrial Ethernet network, with a communication rate set at 1 Gbps and a communication delay strictly controlled within 10 ms to ensure the real-time nature of instruction transmission. Each control unit uses a precise time protocol algorithm for clock synchronization, with a synchronization accuracy better than 1 μs, forming a real-time monitoring and control platform with high synchronization. The platform continuously collects sensor data to achieve comprehensive monitoring of the jacking process. In a certain jacking project, the master control unit processes data streams from 12 jacking points in real time, with a data collection frequency of 100 Hz, and the time deviation of the 12 sub-control units is controlled within 0.8 μs through the clock synchronization algorithm. The assembly module 103 first performs accurate measurement and calibration of the assembly platform, adjusts the support point height using an adjustable support device, and the support point height adjustment accuracy reaches 0.1 mm. Under the guidance of the control network, the steel beam segments are transported to the assembly platform in order, connected with adjacent steel beam segments using high-strength bolts, and the tightening torque is controlled within the range of 1100-1200 N·m. After assembly, a three-dimensional laser scanner is used to scan the steel beam comprehensively, obtain point cloud data, and the point cloud density reaches 1 point / cm². Through a point cloud registration algorithm, the multi-station point cloud data is spliced into a complete digital model of the steel beam. Geometric analysis is performed on the model to extract geometric parameters such as length, width, height, diagonal line, and twist angle of the steel beam, and generate a steel beam geometric parameter dataset. In a certain project, the digital model of a 130 m long large-span steel beam established through three-dimensional laser scanning shows a length error of 3 mm and a twist angle deviation of 0.02° compared with the design value, fully meeting the design accuracy requirements.

[0026] Module 104 is established to optimize the parameters of the hierarchical distributed control architecture and create a neural network prediction model. This module first applies a load to the jacking system using a simulated loading device to simulate the weight of the steel beam and the jacking resistance. In actual operation, the beam weight distribution is calculated based on the geometric parameters of the steel beam, the corresponding load is applied, and the displacement, pressure, and velocity parameters of each jacking point are collected. The Ziegler-Nichols method is used to determine the initial parameters of the PID controller, including the proportional coefficient Kp, the integral coefficient Ki, and the differential coefficient Kd. These parameters are then globally optimized using a genetic algorithm, with the evaluation functions including response time, overshoot, and steady-state error. Simultaneously, a fuzzy control rule library containing 49 rules is constructed based on expert experience, covering different combinations of displacement error and error change rate. An adaptive algorithm is used to dynamically adjust the fuzzy membership function parameters. Finally, the optimized PID parameters, fuzzy control rules, and system response data are input into a BP neural network with 15 hidden layer nodes. Training is performed using the Levenberg-Marquardt algorithm, with the training data including the system response characteristics under different load conditions. This generates a neural network prediction model capable of predicting the system's dynamic response.

[0027] The pushing module 105 performs multi-point synchronous pushing of steel beams based on the neural network prediction model. The module divides the pushing process into four stages: starting, uniform speed, deceleration, and positioning, and sets different speed and acceleration control parameters for each stage. In the starting stage, the speed gradually increases from 0 to 5mm / min, and the acceleration does not exceed 0.2mm / min²; in the uniform speed stage, a constant speed of 10-15mm / min is maintained; in the deceleration stage, the speed drops to 3mm / min, and the deceleration does not exceed 0.3mm / min²; in the positioning stage, the speed is controlled within 1mm / min. The module selects the pushing point at the center position as the master control point and other points as slave control points to construct a master-slave control structure and establish a coordinated control mechanism. The synchronization status of each pushing point is monitored through a real-time monitoring control platform. When the synchronization error is detected to exceed 0.5mm, a feedforward-feedback composite control strategy is used to dynamically adjust the speed and thrust of the corresponding pushing point to ensure the overall synchronous movement of the steel beam. Each pushing point executes a step cycle action, including four action sequences: vertical hydraulic cylinder descending and locking, horizontal hydraulic cylinder advancing, vertical hydraulic cylinder rising and releasing, and horizontal hydraulic cylinder retracting. The pushing distance of each cycle is 800-900mm, realizing precise pushing of steel beams.

[0028] In a specific embodiment, the measurement module 101 is configured to:

[0029] Use a multi-channel geological radar detector to detect the foundation under the jacking track, collect foundation bearing capacity and stability data, and generate a foundation parameter model;

[0030] Based on the foundation parameter model, a control network is established using a precision total station, no less than 8 control points are set and closed measurement is performed to obtain control network data with an error of no more than 1 mm / km;

[0031] Using a digital level to measure the elevation of the control points, and using a closed measurement method to establish an elevation control network;

[0032] According to the control network data and the elevation control network, displacement monitoring points are set every 50m along the jacking track, and settlement monitoring points are set every 30m to construct a real-time deformation monitoring system.

[0033] Specifically, the multi-channel geological radar detector is a device that can non-destructively detect underground structures. It detects the foundation conditions by emitting electromagnetic waves and receiving signals reflected by the underground medium. In the construction of large-span steel beam jacking, this device is used to conduct comprehensive detection of the foundation under the jacking track. During the detection process, the geological radar emits electromagnetic waves at different frequencies (usually 100MHz to 1GHz), penetrates the surface, and generates reflections when encountering the interface between different media. The reflected signal is received by the instrument and converted into a digital signal. The density, uniformity and thickness of the foundation are analyzed through the signal processing algorithm. After the detection data is processed, the foundation bearing capacity data (expressed in kPa) and stability data (usually expressed as a percentage of compaction) are calculated, and combined with the geological parameters to generate a three-dimensional foundation parameter model. The model intuitively displays the foundation bearing capacity distribution and potential weak areas. Based on the foundation parameter model, technicians can determine the optimal layout position of the control network to avoid unstable foundation areas. A precision total station is a high-precision measuring instrument that integrates angle measurement, distance measurement, and data processing. Its angle measurement accuracy typically reaches 1", and its distance measurement accuracy can reach 1mm+1ppm. In actual operation, first, based on the foundation parameter model, stable locations are selected to set no fewer than eight control points. These points should be evenly distributed around the construction area. Subsequently, a closed traverse method is used to connect the control points to form a closed polygonal network. During the measurement process, the total station measures the horizontal angle, vertical angle, and slope distance of the foresight and backsight points at each measuring station. The plane coordinates of each control point are calculated through coordinate calculation and coordinate adjustment. Through a rigorous measurement and calculation process, the closure error of the control network is ensured to not exceed 1mm / km, thereby establishing high-precision control network data.

[0034] Using a digital level to measure the elevation of each control point in the control network is a key step in establishing a height control network. A digital level is a precision instrument that automatically reads and digitally processes level rod markings, typically achieving a measurement accuracy of 0.3 mm / km. The closed leveling route method is used for measurement. Starting from a known elevation point, the elevation of each control point is measured along a predetermined route, ultimately returning to the starting point to form a closed loop. At each measuring station, the level is aimed at a barcoded level rod placed at the control point, automatically reading and storing the elevation difference data. After the measurement is completed, the precise elevation of each control point is determined through adjustment calculations based on the leveling data. Using the closed measurement method effectively controls elevation measurement errors and helps establish a height control network. Based on the established control network data and height control network, technicians deployed monitoring points along the jacking track to establish a real-time deformation monitoring system. Displacement monitoring points are located every 50 meters to monitor horizontal track displacement changes, while settlement monitoring points are located every 30 meters to monitor vertical track deformation. These monitoring points are usually marked with precise markers, such as forced centering monitoring piers or buried monitoring nails, to ensure the stability of the monitoring benchmark. Each monitoring point is equipped with a high-precision displacement sensor or settlement sensor, which can detect tiny displacement changes (usually with an accuracy of 0.01mm) and transmit the data to a data collector. The data collector transmits real-time data to the monitoring control center via an industrial-grade wireless network or wired network, enabling continuous monitoring of track deformation. The monitoring system integrates data analysis software to process and analyze monitoring data in real time, and automatically alarms when abnormal deformation is detected, providing safety protection for jacking construction.

[0035] In a specific embodiment, the configuration module 102 is configured to:

[0036] Configure a high-performance industrial computer as the main control unit, install a real-time operating system, achieve a processing capacity of more than 10GFLOPS, and build a redundant backup system;

[0037] According to the steel beam structural characteristics and weight distribution data in the construction control database, a sub-control unit is configured at each jacking point, and a programmable logic controller is used to execute the jacking instruction;

[0038] Connecting the main control unit to the sub-control unit via industrial Ethernet;

[0039] The main control unit and the sub-control unit are clock synchronized, and a precision time protocol algorithm is used to control the clock synchronization accuracy of all control units to be better than 1μs, forming a real-time monitoring and control platform.

[0040] Specifically, the main control unit is installed with a real-time operating system such as QNX, VxWorks, or RTLinux. These operating systems ensure deterministic and real-time command processing, with system response times typically controlled within milliseconds or even microseconds. The main control unit's processing power must exceed 10 GFLOPS (gigafloating-point operations per second) to meet the demands of complex algorithmic operations and multi-point coordinated control. Furthermore, a redundant backup system, including hot-standby servers, RAID disk arrays, and uninterruptible power supplies, is implemented to ensure seamless failover in the event of a main system failure, preventing data loss and control interruption. Based on the structural characteristics and weight distribution data of the steel beams stored in the construction control database, a sub-control unit is deployed at each jacking point. Steel beam structural characteristics include segment length, cross-sectional characteristics, and connection methods, while weight distribution data accurately records the load conditions at each beam location. The sub-control units utilize programmable logic controllers (PLCs), such as industrial-grade PLCs from the Siemens S7 series, AB ControlLogix series, or Mitsubishi Q series. Each sub-control unit is responsible for executing jacking commands from the main control unit and managing local sensors and actuators. The sub-control unit, equipped with a high-speed processor and real-time operating system, possesses independent computing power and can handle local closed-loop control tasks such as hydraulic cylinder extension and retraction control, pressure regulation, and position feedback. It also provides real-time feedback of execution results and monitoring data to the main control unit, forming a complete control loop. Industrial Ethernet serves as the communication network foundation connecting the main control unit and the sub-control unit. Compared to standard commercial Ethernet, industrial Ethernet offers higher reliability, determinism, and real-time performance, enabling stable operation in industrial sites subject to strong electromagnetic interference and harsh environments. The system utilizes industrial Ethernet protocols such as PROFINET, EtherCAT, or EtherNet / IP, with a communication rate of 1 Gbps or higher to ensure real-time transmission of large amounts of data. The network topology adopts a star or redundant ring structure to enhance the system's fault tolerance. Communication latency is strictly controlled to less than 10ms, ensuring real-time command transmission and data feedback. Network equipment, including industrial-grade switches, fiber optic transceivers, and network isolators, features high protection levels and strong anti-interference capabilities, adapting to the harsh conditions of construction sites.

[0041] Clock synchronization between the master and sub-control units is a key technology for achieving precise, coordinated multi-point control. The system utilizes precision time protocol algorithms, such as IEEE 1588 PTP (Precision Time Protocol), to achieve clock synchronization accuracy of better than 1μs across all control units. In the PTP protocol, the master control unit, acting as the grandmaster clock, periodically broadcasts synchronization messages to the network. The sub-control units, acting as slave clocks, receive these messages and calculate time deviations. By measuring and compensating for network transmission delays, each sub-control unit adjusts its own clock to maintain high synchronization with the master clock. This clock synchronization process is continuous, dynamically compensating for clock drift caused by factors such as temperature fluctuations. High-precision clock synchronization ensures strict timing consistency across all jacking points, laying the foundation for the synchronized jacking of steel beams and ultimately forming a highly synchronized and coordinated real-time monitoring and control platform.

[0042] In a specific embodiment, the assembly module 103 is used to:

[0043] Based on the reference points of the control network, the assembly platform is measured and calibrated, and the height of the support points is adjusted using an adjustable support device to control the height error of the support points to be within 0.1mm;

[0044] Transporting the steel beam segments manufactured in the order of numbering to the assembly platform, and connecting adjacent steel beam segments with high-strength bolts;

[0045] Use a 3D laser scanner to perform a full-scale scan of the assembled steel beams to obtain point cloud data, and then construct a digital model of the steel beams using a point cloud registration algorithm.

[0046] A geometric analysis is performed on the digital model of the steel beam to extract geometric parameters such as the length, width, height, diagonal and torsion angle of the steel beam, thereby generating a geometric parameter data set of the steel beam.

[0047] Specifically, based on the control network benchmark points established in the early stage, a precision total station is used to measure and lay out the assembly platform to determine the theoretical position of the support points. The assembly platform is usually composed of a reinforced concrete structure, and the surface flatness is required to be controlled within 3mm / m. Each support point is installed with an adjustable support device, which consists of a precision screw, an adjustment handwheel and a locking mechanism, with an adjustment accuracy of up to 0.01mm. Based on the measurement results of the total station, the surveyor accurately adjusts the height of each support point to match the design elevation, and the height error is controlled within 0.1mm. The spacing between support points is generally not more than 3m to ensure uniform support for the steel beam. After the adjustment is completed, a precision digital level is used to review the elevation of each support point to confirm that the elevation error meets the requirements, and then all support devices are locked to form a stable assembly foundation.

[0048] After the steel beam segments are manufactured, they are transported to the assembly platform in numerical order for assembly. Steel beam segments are typically manufactured in-house from Q345qD high-strength low-alloy structural steel. Each segment is typically 12-15 meters long for easy transportation. Each segment is uniquely numbered according to the design drawings to ensure the correct assembly sequence. Dedicated transport vehicles are used during transportation to prevent deformation of the beam. Upon arrival at the site, crane equipment is used to precisely position the beam segments on the assembly platform's support points. Adjacent beam segments are connected using high-strength bolts, typically M27, made of grade 10.9 high-strength bolts with a tensile strength exceeding 1000 MPa. The connecting plate is made of the same steel material as the main beam, with a thickness no less than that of the main beam web. During installation, the connecting plate is temporarily secured to ensure alignment of the holes. High-strength bolts are then installed and tightened to the specified torque (1100-1200 N·m) using a torque wrench to ensure a secure connection. After assembly, a 3D laser scanner is used to scan the entire beam in all directions, acquiring precise point cloud data. A 3D laser scanner is a measuring device that rapidly acquires three-dimensional coordinate data on an object's surface, typically with an accuracy of ±2mm / 100m and a scanning speed of up to 1 million points / second. Due to the large size of long-span beams, scanning is performed at multiple stations to cover the entire beam surface. Each scanning station is positioned based on a control network to ensure coordinate consistency across all stations. During the scanning process, a laser beam illuminates the beam surface from different angles. The reflected light signals are received by the instrument and converted into 3D coordinate data, forming a dense point cloud dataset. The point cloud density is typically set to 1 point / cm² to ensure sufficient detail. The raw point cloud data is then filtered to remove environmental interference and outliers. A point cloud registration algorithm is then used to stitch the multi-station cloud data into a complete beam point cloud model. This registration algorithm typically employs an iterative closest point (ICP) algorithm or a feature matching algorithm. This algorithm identifies overlapping areas in adjacent point clouds and calculates a transformation matrix to achieve precise point cloud stitching.

[0049] Geometric analysis is performed on the point cloud data of the assembled steel beam to extract key geometric parameters. This process first converts the point cloud data into a 3D mesh model or NURBS surface model to facilitate geometric feature extraction. A plane fitting algorithm is then used to identify the main structural surfaces of the steel beam, such as the upper and lower flanges, and the web. Based on these identified structural surfaces, the geometric parameters of the steel beam are calculated: length is measured by extracting the distance between the two end points; width is obtained by measuring the distance between the outer edges of the flanges; height is calculated by the perpendicular distance between the centerlines of the upper and lower flanges; diagonal is determined by measuring the spatial distance between the four corner points of the steel beam; and torsion angle is calculated by analyzing the angular variation between the normal vector of each section's flange surface and a reference plane. These geometric parameters are compared with the design values, and deviations are calculated to assess assembly quality. All extracted geometric parameters and their deviations are organized into a structured dataset, the steel beam geometric parameter dataset, which serves as an important basis for subsequent jacking control. This dataset contains the global dimensional parameters of the steel beam and precise measurements of key local locations, providing data support for precise control during the jacking process.

[0050] In a specific embodiment, the establishing module 104 is configured to:

[0051] Applying a load to the jacking system through a simulated loading device to simulate the weight of the steel beam and the jacking resistance corresponding to the geometric parameters of the steel beam, and collecting the displacement, pressure and velocity parameters of each jacking point;

[0052] Based on the data collected by the real-time monitoring and control platform, the initial parameters of the PID controller are determined using the Ziegler-Nichols method, and then globally optimized using a genetic algorithm;

[0053] A fuzzy control rule base is constructed based on expert experience and the geometric parameters of the steel beam, which contains 49 rules, and the fuzzy membership function parameters are dynamically adjusted through an adaptive algorithm;

[0054] The initial parameters of the PID controller, fuzzy control rules and system response data are input into a BP neural network, which is trained using a Levenberg-Marquardt algorithm to generate a neural network prediction model.

[0055] Specifically, in a multi-point synchronous walking jacking construction system for large-span steel beams without piers, parameter optimization of the control system requires a simulated loading test. The simulated loading device is a specialized device capable of simulating the actual load characteristics of a steel beam. It consists of a hydraulic loading system, load sensors, and a control system. Based on data from the beam's geometric parameter dataset, the weight distribution of each beam component and the expected resistance during jacking are calculated. For example, a beam with a span of 120 meters weighs approximately 450 tons. Its geometric parameters allow the center of gravity and the theoretical load at each jacking point to be calculated. The simulated loading device applies loads to the jacking system using multiple hydraulic cylinders according to the calculated load distribution, simulating the beam's weight and jacking resistance. During the loading process, displacement sensors, pressure sensors, and velocity sensors installed at each jacking point collect data in real time. The displacement sensors have an accuracy of 0.01 mm, the pressure sensors have an accuracy of 0.1% FS (full scale), and the velocity sensors have an accuracy of 0.1 mm / min. These sensor data are transmitted to a real-time monitoring and control platform via a data acquisition system at a frequency of 100 Hz to ensure that the dynamic characteristics of the system are captured. The test process simulates a complete jacking cycle, including starting, constant speed operation, speed change and precise positioning, and comprehensively collects the response characteristics of the system under various working conditions.

[0056] Based on data collected by the real-time monitoring and control platform, the system uses the Ziegler-Nichols method to determine the initial parameters of the PID controller. The Ziegler-Nichols method is a classic PID parameter tuning method, which is divided into the critical oscillation method and the step response method. In push-pushing systems, the critical oscillation method is commonly used: first, the integral time Ti is set to infinity (i.e., Ki = 0) and the differential time Td is set to 0 (i.e., Kd = 0). Then, the proportional gain Kp is gradually increased until the system exhibits constant-amplitude oscillations. The proportional gain at this point is recorded as the critical gain Ku, and the oscillation period is Tu. According to the Ziegler-Nichols tuning formula, the initial PID parameters are set as: Kp = 0.6Ku, Ti = 0.5Tu, and Td = 0.125Tu. However, the parameters obtained by this method are often only a rough estimate and require further optimization. The system introduces a genetic algorithm for global optimization of the PID parameters. The genetic algorithm searches for the optimal solution by simulating natural selection and genetic mechanisms. The algorithm first generates an initial population of 50 individuals based on the parameters obtained by the Ziegler-Nichols method, with each individual representing a set of PID parameters (Kp, Ki, Kd). The evaluation function comprehensively considers system response time, overshoot, and steady-state error. A new generation of populations is generated through genetic operations such as crossover and mutation, and the parameters are iteratively optimized. After approximately 100 generations of iterative calculations, the optimized PID parameters are obtained, ensuring a system response time of no more than 100 ms, overshoot within 2%, and steady-state error within 0.1 mm.

[0057] To address the nonlinear characteristics and uncertainties of the steel beam pushing process, the system constructs a fuzzy control rule base based on expert experience and beam geometric parameters. Fuzzy control is a control method based on fuzzy set theory and fuzzy logic that can simulate the decision-making process of human experts. When building the fuzzy rule base, the input and output linguistic variables are first defined. The input variables typically include the displacement error e and the error change rate ec, and the output variable is the control increment u. Each variable is divided into several fuzzy sets, such as negative large (NB), negative medium (NM), negative small (NS), zero (ZO), positive small (PS), positive medium (PM), and positive large (PB). Triangular or Gaussian functions are used as membership functions. The fuzzy rule base contains 49 if-then rules, covering 7×7 different combinations of displacement error and error change rate. For example, "IF e is NB AND ec is NB THEN u is PB" means that when the displacement error is negative and the error change rate is negative, the control increment should be positive. The system also uses an adaptive algorithm to dynamically adjust the parameters of the fuzzy membership functions. The algorithm analyzes the system response data, evaluates the effect of the current fuzzy control, and automatically adjusts the center point and width parameters of the membership function according to the performance indicators, so that the fuzzy control rules can adapt to the changes in system characteristics and improve the control accuracy.

[0058] The optimized PID controller parameters, fuzzy control rules, and system response data are used as training samples and fed into a BP neural network for training. This generates a neural network model capable of predicting the system's dynamic characteristics. The BP neural network is a multi-layer feedforward neural network that uses a backpropagation algorithm to adjust network weights and possesses strong nonlinear mapping capabilities. The network structure is designed to consist of three layers: input, hidden, and output. The input layer nodes contain parameters such as the current position, velocity, and applied load of each push point. The hidden layer has 15 nodes and uses a sigmoid activation function. The output layer nodes correspond to the predicted system responses, such as the displacement, velocity, and acceleration of each push point. The BP neural network is trained using the Levenberg-Marquardt algorithm, which combines the advantages of the steepest descent method and the Gauss-Newton method, resulting in rapid convergence and high training accuracy. The training dataset consists of system response characteristics recorded under different load conditions and control parameters, typically with a sample size of at least 1,000. During training, the network weights are gradually adjusted to minimize the mean squared error between the network output and the target output. Training ends when the mean square error drops below a preset threshold (typically 0.001) or reaches the maximum number of iterations (typically 1,000). The trained neural network can accurately predict the system's dynamic response to a given control input, with prediction errors within 1%, providing strong support for achieving high-precision, multi-point synchronous thrusting.

[0059] In a specific embodiment, the pushing module 105 is used to:

[0060] According to the neural network prediction model, the steel beam pushing process is divided into the starting stage, the constant speed stage, the deceleration stage and the positioning stage, and the speed and acceleration control strategies for each stage are formulated;

[0061] The central pushing point is selected as the master control point, and other points are selected as slave control points. A master-slave control structure is constructed to establish a coordinated control mechanism between the pushing points.

[0062] The synchronization status of each pushing point is monitored by the real-time monitoring and control platform. When the synchronization error is detected to be greater than 0.5 mm, the speed and thrust of the corresponding pushing point are adjusted by a feedforward-feedback composite control strategy.

[0063] A step cycle action is executed for each pushing point, including four action sequences: vertical hydraulic cylinder descending and locking, horizontal hydraulic cylinder advancing, vertical hydraulic cylinder rising and releasing, and horizontal hydraulic cylinder retracting, to achieve synchronous movement of the steel beam.

[0064] Specifically, based on the analysis results of a neural network prediction model, the system scientifically divides the steel beam pushing process into four phases, developing precise control strategies for each phase. The initial phase utilizes a low-speed start-up strategy, gradually increasing the speed from 0 to 5 mm / min, with acceleration strictly controlled within 0.2 mm / min², and lasting for at least 10 minutes. This gradual start-up approach avoids vibration and deformation of the steel beam that could be caused by sudden application of thrust. The constant-speed phase, the main part of the pushing process, accounts for over 70% of the total pushing distance. The speed is maintained at a constant value of 10-15 mm / min, a speed range that ensures both construction efficiency and system stability. The deceleration phase begins before approaching the target position, gradually reducing the speed from a constant speed to 3 mm / min, with deceleration controlled within 0.3 mm / min², and lasting for at least 8 minutes to ensure a smooth system transition. The positioning phase, the final stage of the pushing process, utilizes a precise control strategy, maintaining the speed within 1 mm / min, until the steel beam reaches the designed position, with positioning accuracy within ±2 mm. This phased control strategy fully considers the system's dynamic characteristics, ensuring smooth and precise jacking. To achieve multi-point synchronous jacking, the system employs a master-slave control structure, selecting the central jacking point as the master control point and the remaining points as slaves. The master control point is typically located near the center of gravity of the beam, where its load conditions are most typical and represent the overall jacking state. The system assigns a reasonable load ratio to each jacking point based on the beam's geometric characteristics and weight distribution. In the control algorithm, the master control point's trajectory serves as the reference trajectory, and its displacement, velocity, and acceleration form the basis for control commands. Slave control points adjust control parameters based on the master control point's motion state and their own positional deviations to ensure synchronized movement with the master control point. The system also establishes a coordinated control mechanism among the jacking points. Through a distributed control algorithm, these points coordinate with each other to adapt to the overall motion requirements of the beam. This coordination mechanism, which includes load balancing, synchronization error compensation, and abnormal state handling algorithms, ensures uniform load distribution on the beam during jacking and avoids local overload or underload.

[0065] The real-time monitoring and control platform continuously monitors the synchronization status of each push point, including parameters such as displacement, velocity, pressure, and inclination. Each push point is equipped with a high-precision displacement sensor to monitor real-time displacement and calculate the synchronization error relative to the master control point. If the synchronization error detected at a push point exceeds the set threshold of 0.5mm, the system immediately initiates an adjustment mechanism. This adjustment process utilizes a feedforward-feedback composite control strategy. Feedforward control, based on a neural network prediction model, predicts the system's future response based on the current state and pre-calculates the required control input. Feedback control, based on the actual monitored synchronization error, makes real-time corrections to the control input. These two control approaches complement each other, with feedforward control providing rapid response and feedback control ensuring control accuracy. If the synchronization error exceeds the threshold, the system dynamically adjusts the speed and thrust of the corresponding push point based on the error magnitude and trend. The speed adjustment range is ±20% of the current speed, and the thrust adjustment range is ±15% of the current thrust. This precise adjustment ensures high synchronization among all push points, with the synchronization error consistently within a ±0.5mm range.

[0066] The walking mechanism at each pushing point performs a specific sequence of movements to achieve cyclical advancement of the beam. This sequence consists of four sequential steps: First, the vertical hydraulic cylinder descends to lock the beam. The vertical cylinder descends at a speed of 5-8 mm / s, applying a downward pressure of at least 300 tons to ensure firm contact and lock with the bottom of the beam. Next, the horizontal hydraulic cylinder advances, moving forward 800-900 mm at the set pushing speed. The thrust is calculated based on the load distribution, typically 200-300 tons. Next, the vertical hydraulic cylinder ascends to release the beam at a speed of 5-8 mm / s, with a lift of at least 50 mm, completely clearing the bottom of the beam. Finally, the horizontal hydraulic cylinder retracts at a speed of 15-20 mm / s, preparing for the next advancement cycle. These four movements constitute a complete walking cycle, with a time limit of 3-4 minutes. The walking cycles at each pushing point are highly synchronized, with a phase difference of less than 0.1 second, ensuring synchronized movement of the entire beam. The coordinated work of multiple jacking points forms a distributed propulsion system, which can evenly bear the weight of the steel beam and the jacking resistance, avoid local stress concentration, and achieve smooth and precise jacking of large-span steel beams, providing reliable technical support for pier-free jacking construction.

[0067] In a specific embodiment, the correction module 106 is configured to:

[0068] Based on the control network, a high-precision total station is used to set up no less than 50 measurement points at key locations of the steel beam, collect three-dimensional coordinate data, and establish a model of the actual position of the steel beam;

[0069] Comparing and analyzing the actual position model of the steel beam with the design parameters in the construction control database, and calculating the plane position deviation, elevation deviation, longitudinal deviation and lateral deviation;

[0070] According to the calculated position deviation, a precision adjustment system consisting of a precision jack and a wedge-shaped pad is used for fine-tuning, and the control adjustment accuracy reaches 0.1mm;

[0071] Use a precision level to re-measure the adjusted steel beam to verify that the plane position deviation is controlled within ±5mm and the elevation deviation is controlled within ±3mm.

[0072] Specifically, in the multi-point synchronous walking jacking construction system for large-span steel beams without piers, the precise measurement and correction of the steel beams after they are pushed into place are the key links to ensure the quality of the project. Based on the control network established in the early stage, technicians use high-precision total stations to conduct comprehensive measurements of the steel beams that have been pushed into place. The total station is a high-precision measuring instrument that integrates angle measurement and distance measurement. Its angle measurement accuracy usually reaches 1", and its distance measurement accuracy can reach 1mm+1ppm. No less than 50 measuring points are arranged at key positions of the steel beams. These points include important locations such as support positions, mid-span positions, quarter-point positions, and structural changes. The measuring points collect three-dimensional coordinate data (X, Y, Z). To ensure measurement accuracy, the total station is set up on a known point in the control network, and a coordinate system is established by looking back at other known points to eliminate instrument errors. The measurement data is transmitted to the computer in real time through the data transmission interface, and processed using professional measurement software to generate a point cloud model of the steel beam. Through the three-dimensional modeling algorithm, discrete measurement points are connected into a surface model to form a complete model of the actual position of the steel beam, which accurately reflects the spatial position status of the steel beam. The actual position model of the steel beam is compared and analyzed with the design parameters stored in the construction control database to calculate various position deviations. The construction control database contains precise data such as the design position coordinates, elevation, and axis direction of the steel beam. The comparison analysis is performed using dedicated bridge measurement and analysis software, which can automatically calculate the measured model and the design model. The plane position deviation refers to the offset of the center line of the steel beam relative to the design position in the horizontal plane, which is obtained by calculating the Euclidean distance between the measured coordinates and the design coordinates in the XY plane; the elevation deviation refers to the deviation of the steel beam in the vertical direction relative to the design elevation, which is calculated by the difference between the measured Z coordinate and the design Z coordinate; the longitudinal deviation refers to the position deviation of the steel beam in the direction of the bridge axis, which is obtained by projecting the measured coordinates onto the design axis and calculating the distance between the projection point and the design position; the lateral deviation refers to the position deviation of the steel beam in the direction perpendicular to the bridge axis, which is obtained by calculating the shortest distance from the measured point to the design axis. The analysis software generates a deviation distribution diagram and a deviation statistical report, which intuitively displays the deviation of each part and marks the out-of-tolerance points to provide a basis for subsequent precise adjustments.

[0073] Based on the calculated position deviation data, technicians use a precision adjustment system to fine-tune the position of the steel beam. The precision adjustment system is mainly composed of precision jacks and wedge-shaped pads. The precision jack is a hydraulic device that can achieve precise lifting and lowering. It has a load-bearing capacity of not less than 200 tons, a stroke of generally 100-150mm, and an adjustment accuracy of up to 0.1mm. The precision jack is equipped with high-precision pressure sensors and displacement sensors to monitor the lifting force and lifting height in real time. The wedge-shaped pad is a precision wedge-shaped block made of high-strength alloy steel. It is used in pairs and can achieve fine-tuning of height through relative sliding. The surface hardness of the wedge-shaped pad is not less than HRC45, and the surface roughness Ra is not greater than 1.6μm, ensuring smooth sliding and precise positioning. The adjustment process begins by lifting the steel beam with a precision jack at a slow rate of 0.5-1mm / min to avoid impact loads. Then, based on the calculated deviation, wedge-shaped spacers of appropriate thickness are inserted or their relative positions are adjusted to achieve fine-tuning in the horizontal and vertical directions. Finally, the jack is slowly released to stabilize the beam in its adjusted position. The entire adjustment process is conducted under real-time monitoring by a measurement system, ensuring an adjustment accuracy of 0.1mm.

[0074] After the adjustments were completed, the steel beams were re-measured using a precision level and total station to verify the effectiveness of the adjustments. A precision level, a precision instrument designed specifically for elevation measurement, boasts an accuracy of up to 0.3mm / km and is used to precisely measure the elevation of the steel beams. The re-measurement used the same measurement point layout as the initial measurement to ensure data comparability. The re-measured data was then compared and analyzed with the design data to verify that all deviations met design requirements. According to the technical specifications for bridge engineering, the planar position deviation of long-span steel beams should be controlled within ±5mm, the elevation deviation within ±3mm, the longitudinal deviation within ±8mm, and the lateral deviation within ±4mm. If the re-measurement results indicate that certain areas still exceed the allowable deviation, the adjustment and re-measurement process must be repeated until all deviation indicators meet the requirements. Once verified, permanent supports will be installed and the temporary support system removed. Complete measurement and adjustment data will be recorded in the project archives, serving as the basis for quality inspection and providing essential data for subsequent bridge maintenance. Through this series of precise measurements and adjustments, we ensure that the final positioning accuracy of the large-span steel beams meets the design requirements, laying the foundation for the safe operation of the bridge.

[0075] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A large-span steel beam without piers multi-point synchronous walking type jacking construction system, characterized in that the system include: The measurement module is used to measure the construction site, establish a control network and elevation control network, arrange displacement monitoring points, and generate a construction control database; A configuration module is used to configure a main control unit and sub-control units based on the construction control database and adopt a hierarchical distributed control architecture to form a real-time monitoring and control platform. Specifically, the module is used to: configure a high-performance industrial computer as the main control unit, install a real-time operating system, and achieve a processing capacity of more than 10GFLOPS to build a redundant backup system; configure a sub-control unit at each jacking point based on the steel beam structural characteristics and weight distribution data in the construction control database, and use a programmable logic controller to execute jacking instructions; connect the main control unit with the sub-control units via industrial Ethernet; synchronize the clocks of the main control unit and the sub-control units, and use a precision time protocol algorithm to control the clock synchronization accuracy of all control units to be better than 1μs, thereby forming a real-time monitoring and control platform; An assembly module is used to assemble steel beams in a numbered sequence under the guidance of the control network, establish a digital model of the steel beam using three-dimensional laser scanning, and determine the geometric parameters of the steel beam. Specifically, it is used to: measure and calibrate the assembly platform based on the reference points of the control network, adjust the support point height using an adjustable support device, and control the support point height error to within 0.1mm; transport the steel beam segments manufactured in a numbered sequence to the assembly platform, and connect adjacent steel beam segments using high-strength bolts; use a three-dimensional laser scanner to perform an omnidirectional scan of the assembled steel beam to obtain point cloud data, and construct a digital model of the steel beam using a point cloud registration algorithm; perform geometric analysis on the digital model of the steel beam, extract the length, width, height, diagonal and torsion angle of the steel beam, and generate a data set of geometric parameters of the steel beam; Establish a module for performing parameter optimization on the hierarchical distributed control architecture, and establish a neural network prediction model based on the geometric parameters of the steel beam and the data collected by the real-time monitoring and control platform, specifically for: applying a load to the jacking system by simulating a loading device, simulating the weight of the steel beam and the jacking resistance corresponding to the geometric parameters of the steel beam, and collecting the displacement, pressure, and velocity parameters of each jacking point; determining the initial parameters of the PID controller using the Ziegler-Nichols method based on the data collected by the real-time monitoring and control platform, and then performing global optimization using a genetic algorithm; constructing a fuzzy control rule library containing 49 rules based on expert experience and the geometric parameters of the steel beam, and dynamically adjusting the parameters of the fuzzy membership function through an adaptive algorithm; inputting the initial parameters of the PID controller, the fuzzy control rules, and the system response data into a BP neural network, training it using the Levenberg-Marquardt algorithm, and generating a neural network prediction model; A pushing module is used to perform multi-point synchronous pushing of steel beams based on the neural network prediction model, adopt a feedforward-feedback composite control strategy, and dynamically adjust the parameters of each pushing point through the real-time monitoring and control platform; The correction module is used to measure the spatial position of the steel beam after the jacking is completed using the control network, calculate the position deviation according to the design parameters in the construction control database, and perform position correction according to the position deviation.

2. The large-span steel beam pier-free multi-point synchronous walking jacking construction system according to claim 1 is characterized in that: The measuring module is used for: Use a multi-channel geological radar detector to detect the foundation under the jacking track, collect foundation bearing capacity and stability data, and generate a foundation parameter model; Based on the foundation parameter model, a control network is established using a precision total station, no less than 8 control points are set and closed measurement is performed to obtain control network data with an error of no more than 1 mm / km; Using a digital level to measure the elevation of the control points, and using a closed measurement method to establish an elevation control network; According to the control network data and the elevation control network, displacement monitoring points are set every 50m along the jacking track, and settlement monitoring points are set every 30m to construct a real-time deformation monitoring system.

3. The large-span steel beam pier-free multi-point synchronous walking jacking construction system according to claim 1 is characterized in that: The pushing module is used to: According to the neural network prediction model, the steel beam pushing process is divided into the starting stage, the constant speed stage, the deceleration stage and the positioning stage, and the speed and acceleration control strategies for each stage are formulated; The central pushing point is selected as the master control point, and other points are selected as slave control points. A master-slave control structure is constructed to establish a coordinated control mechanism between the pushing points. The synchronization status of each pushing point is monitored by the real-time monitoring and control platform. When the synchronization error is detected to be greater than 0.5 mm, the speed and thrust of the corresponding pushing point are adjusted by a feedforward-feedback composite control strategy. A step cycle action is executed for each pushing point, including four action sequences: vertical hydraulic cylinder descending and locking, horizontal hydraulic cylinder advancing, vertical hydraulic cylinder rising and releasing, and horizontal hydraulic cylinder retracting, to achieve synchronous movement of the steel beam.

4. The large-span steel beam pier-free multi-point synchronous walking jacking construction system according to claim 3 is characterized in that: The correction module is used to: Based on the control network, a high-precision total station is used to set up no less than 50 measurement points at key locations of the steel beam, collect three-dimensional coordinate data, and establish a model of the actual position of the steel beam; Comparing and analyzing the actual position model of the steel beam with the design parameters in the construction control database, and calculating the plane position deviation, elevation deviation, longitudinal deviation and lateral deviation; According to the calculated position deviation, a precision adjustment system consisting of a precision jack and a wedge-shaped pad is used for fine-tuning, and the control adjustment accuracy reaches 0.1mm; Use a precision level to re-measure the adjusted steel beam to verify that the plane position deviation is controlled within ±5mm and the elevation deviation is controlled within ±3mm.

Images