A steel structure construction precision control system based on multi-data fusion

Abstract

The application discloses a steel structure construction precision control system based on multi-data fusion, and particularly relates to the technical field of sensor data fusion, which comprehensively collects component initial processing size, real-time spatial displacement and installation environment data through an integrated sensing network, and based on the three types of data, respectively analyzes initial deviation rate, process deviation rate and environmental deviation rate; an intelligent prediction module predicts one-time pose deviation in the component installation stage; an intelligent decision module constructs a deviation quantization model according to the prediction result and generates an operable pose correction decision; an execution and control module receives the decision instruction, drives an execution mechanism to complete actual adjustment operation, and feeds back the execution state in real time; a closed-loop verification and self-learning module performs secondary pose measurement on the adjusted component, feeds back the data to the model for incremental training, and finally realizes autonomous evolution and continuous improvement of the construction precision control system.

Description

Technical Field

[0001] This invention relates to the field of multi-sensor data fusion technology, and more specifically, to a steel structure construction accuracy control system based on multi-data fusion. Background Technology

[0002] Steel structure construction commonly utilizes automated total stations and digital levels. These two instruments typically work together to form a high-precision measurement and control system for the project. The total station is responsible for setting out and laying out three-dimensional spatial coordinates, while the digital level is responsible for precise elevation transfer and settlement monitoring. Taking steel column installation as an example, the installation process involves hoisting the steel column, aligning the holes in its base plate with the anchor bolts, initially positioning it, and temporarily tightening the nuts. Reflecting prisms are then attached or installed on the upper part of the steel column in two directions at 90 degrees to each other. The prism positions should be calculated so that their coordinates accurately reflect the center and orientation of the steel column. A total station is set up at a control point, and another control point is backsighted for orientation. The total station compares the measured prism coordinates with the design coordinates of that point in the BIM model in real time. The software automatically calculates the deviations of the steel column in the X and Y directions, as well as the verticality deviation. Based on the displayed deviation data, construction personnel use tools to precisely adjust, verify, and fix the steel column. After adjustment, the prism coordinates are measured again with the total station to confirm that the deviation values ​​meet the specifications.

[0003] However, in actual use, it still has some shortcomings: (1) Data chain is fragmented and lacks full-process closed-loop management. The shortcomings are: data collection is scattered and discontinuous, and it is impossible to accurately locate which link or which process introduced the deviation; (2) Data utilization is superficial and lacks intelligent integration and traceability capabilities. Data processing is limited to simple recording and comparison. Multi-source data exists in isolation and fails to be deeply integrated and correlated; (3) Control mode is passive and lacks prediction and self-optimization capabilities. The control system lacks foresight, cannot predict future trends, and can only passively correct deviations after they occur. It cannot dynamically optimize and adjust according to the real-time working conditions on site, and it does not have the ability to learn and evolve from historical data. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art, this invention provides a steel structure construction accuracy control system based on multi-data fusion, which solves the problems mentioned in the background art through the following scheme. It includes a full-state perception module: using an integrated sensor network to comprehensively collect component construction accuracy data, specifically initial processing dimension data of the component collected by a 3D laser scanner, real-time spatial displacement data of the component collected by a GNSS receiver, tilt sensor, and force sensor, and component installation environment impact data collected by a vibration sensor;

[0005] Deviation Calculation Module: Based on the initial machining dimension data of the component, the real-time spatial displacement data of the component, and the influence data of the component installation environment, the initial deviation rate δ of the component installation process is calculated. i Process deviation rate δ p and environmental deviation rate δ e ;

[0006] Initial deviation rate δ i The calculation method is as follows: Where La is the component feature size extracted from the component point cloud data, and Lf is the corresponding feature size extracted from the point cloud data of the proposed installation location;

[0007] Process deviation rate δ p The calculation method is the product of the deviation rate of the component's resultant velocity, the deviation rate of its attitude angle, and the deviation rate of its overall force.

[0008] Environmental Deviation Rate δ e The calculation method is as follows: when the measured vibration frequency f of the component is... m Greater than the safety threshold f a hour, When f m Less than or equal to the safety threshold f a At that time, δ e =0;

[0009] Intelligent prediction module: Input initial deviation rate δ i Process deviation rate δ p and environmental deviation rate δ e The Kalman filter algorithm is used to predict the first pose deviation and total deviation rate of the component. The first pose deviation is a six-degree-of-freedom state vector that characterizes the deviation of the component from the intended installation position. The state vector includes translation deviations of the x-axis, y-axis, and z-axis, as well as rotation angle deviations θx, θy, and θz around the x-axis, y-axis, and z-axis.

[0010] Intelligent Decision Module: Based on the predicted pose deviation results of the component, a deviation quantification model is constructed, and a pose correction decision is generated. The deviation quantification model is used to analyze the influence weight of each deviation rate on the total deviation, and the model form is ΔX. k ≈β i *δ i +β p *δ p +β e *δ e +ε, where β i ,β p ,β e Let δ be the regression coefficient obtained by fitting using the least squares method. i δ p δ eThese are the initial deviation rate, process deviation rate, and environmental deviation rate, respectively, and ε is the residual term, representing the portion that the model could not explain.

[0011] Execution and Control Module: Receives pose correction decisions, drives the execution device to complete the specific operations of the decisions, and uses built-in sensors to provide feedback on the completion status of the decision operations;

[0012] Closed-loop verification and self-learning module: For installed components that have completed the decision-making operation, pose information is collected again, secondary pose deviations are identified, and the secondary pose deviation data is fed back for incremental learning to continuously optimize the pose deviation prediction algorithm and the pose deviation optimization decision algorithm.

[0013] The technical effects and advantages of this invention are as follows:

[0014] 1. Full-process, high-precision data closed loop: This invention constructs a precision data chain covering the entire lifecycle of component installation. This data chain systematically defines and collects key precision data from before, during, and after installation, and quantifies and correlates them through unified standards, forming a complete, consistent, and high-precision data closed loop;

[0015] 2. Multi-source data fusion and intelligent deviation tracing: This invention achieves effective integration and in-depth analysis of data from multiple sources and of different types, identifies various key factors affecting installation accuracy and their relative importance, and realizes the transformation from simple data measurement to intelligent cause tracing;

[0016] 3. Predictive control and continuous self-optimization: An intelligent closed-loop control system has been established. This system can proactively predict possible deviations in the future based on the current state and generate and execute optimized adjustment decisions in advance. The system will feed back the actual effect of each adjustment to itself for continuous learning and improvement. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall structure of the present invention.

[0018] Figure 2 This is a schematic diagram of the execution flow of the present invention. Detailed Implementation

[0019] 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.

[0020] like Figure 1The steel structure construction accuracy control system shown includes a full-state perception module, a deviation calculation module, an intelligent prediction module, an intelligent decision-making module, an execution and control module, and a closed-loop verification and self-learning module.

[0021] like Figure 2 The invention illustrates a steel structure construction accuracy control system based on multi-data fusion, which collects data during component installation, calculates the deviation rate, makes a posture correction decision based on the deviation rate, and collects data on the component again after the decision is executed.

[0022] Full-state perception module: Using an integrated sensor network, it comprehensively collects component construction accuracy data, including initial processing dimension data, real-time spatial displacement data, and environmental impact data of component installation.

[0023] The initial dimensional data of the components are obtained by scanning with a 3D laser scanner. Before the components arrive on site or are hoisted, the dimensional data of the components to be installed are acquired and compared with the dimensional data of the intended installation location to calculate the initial deviation rate δ. i .

[0024] A 3D laser scanner is used to perform a comprehensive scan of the component and the proposed installation location. It is important to note that the scan points must be sufficiently dense to cover all feature dimensions, and there should be no significant obstruction of these dimensions. In this example, the selected scan point spacing is 1mm. Feature dimensions refer to the digital representation of the geometric attributes of the component and the proposed installation location and their spatial relationship within the overall structure. A unified field control network coordinate system is established. Based on the 3D laser scanner data, precise point cloud data of the component and the proposed installation location are obtained. Through point cloud data processing and coordinate system registration, the point cloud data of the component and the proposed installation location are unified to the control network coordinate system, achieving feature dimension alignment. After outputting the unified coordinate system of the feature dimension coordinates of the component and the proposed installation location, initial deviation calculation is performed. For each feature dimension i to be checked, the initial deviation rate calculation formula is: δ i L represents the initial deviation rate of the i-th feature size; a L represents the dimension of the i-th component feature extracted from the component point cloud data, in mm. f The feature dimension of the proposed installation location, extracted from the point cloud data and aligned with the component's feature dimensions; if δ i If δ > 0, then the actual size of the component is larger than the size of the intended installation location; if δ i If the value is less than 0, then the actual size of the component is smaller than the size of the intended installation location.

[0025] For the initial deviation rate pass / fail judgment: the calculated δ i Compare with the deviation tolerance δ0 required by the specification or design, if δ iIf δ < 0, then the dimension is acceptable; if δ i If the value is greater than δ0, then the dimension is out of tolerance and needs to be recorded.

[0026] If the initial deviation rate of all feature dimensions is within the deviation tolerance, the component installation proceeds normally. If the initial deviation rate of some feature dimensions exceeds the deviation tolerance and on-site correction is possible, the feature dimensions are corrected on-site, and the component installation proceeds normally after marking. If the initial deviation rate of some feature dimensions exceeds the deviation tolerance and on-site correction is not possible, the component is returned to the factory for repair.

[0027] Real-time spatial displacement data of components is obtained through GNSS receivers, tilt sensors, and force sensors to monitor the spatial position, attitude, and stress state of components during installation and calculate the process deviation rate δ. p .

[0028] The GNSS receiver calculates the component's x-axis, y-axis, and z-axis velocities and the three-dimensional composite velocity at the endpoint of the field control network coordinate system in real time, tracking the component's motion trajectory; the tilt sensor monitors the component's pitch angle in real time to determine the component's motion stability, identify swaying, and verify whether the installation angle meets the design requirements after positioning; the force sensor collects the load data of each hook in real time, calculates the percentage difference in force between lifting points, and ensures that the load deviation of each point from the average value does not exceed ±10%, thereby avoiding single-point overload or underload.

[0029] The GNSS receiver provides the three-dimensional motion velocity of the component during installation in the field control network coordinate system: including the component's velocity components V along the X, Y, and Z axes at time t. x (t), V y (t), V z (t), and the resultant velocity V(t) representing the instantaneous velocity, are compared to see if the velocities in each direction exceed the safe speed limit, V xa V ya V za and V a These represent the maximum permissible speed and resultant speed in the x, y, and z directions, respectively; calculate the speed deviation rate, taking the x-axis direction as an example, the x-axis speed deviation rate is... The same method is used to calculate the speed deviation rates of the y-axis and z-axis and the total speed deviation rate respectively; if any speed deviation rate exceeds the maximum allowable speed in that direction, the installation is suspended, otherwise the component installation continues.

[0030] The tilt sensor is directly mounted on the component, and its measurements are based on the field control network coordinate system, reflecting the component's attitude relative to the ground horizontal reference, regardless of how the component moves within the coordinate system. The tilt sensor obtains the pitch angle θ, which is compared with the maximum allowable tilt angle θ0. If θ < θ0, the angular deviation rate δ θ=1, not included in the process deviation rate calculation; if θ>θ0, then .

[0031] Force sensor monitors F i (t), that is, the force on the i-th support point during transportation at time t during the component installation process, i=1,2,3,...,n, then the average load is: Force deviation at each support point Overall force deviation rate δF is required for all points i Both δF(t) and δF(t) are less than or equal to 10%.

[0032] Process deviation rate δ p The product of the resultant velocity deviation rate, angular deviation rate, and overall force deviation rate is worth noting. It is important to note that the deviation rate calculation requires the resultant velocity deviation rate, angular deviation rate, and overall force deviation rate to all meet their respective limits.

[0033] Environmental impact data is collected by vibration sensors, which monitor the vibration frequency of components in real time. The vibration of components is caused by the instantaneous kinetic energy generated by adjacent installed machines. The environmental deviation rate δ is calculated based on the frequency. e .

[0034] When f m >f a hour, When f m ≤f a At that time, δ e =0; where f m To measure the vibration frequency of the component, a high-sensitivity vibration sensor collects data in real time, transmitting instantaneous kinetic energy intensity during the operation of adjacent equipment, thus determining the vibration frequency generated by the component; f a The safety threshold is set according to the component. Exceeding this limit may cause component displacement, loose connection or installation deviation.

[0035] The intelligent prediction module predicts the first-order pose deviation and total deviation rate of a component based on the initial deviation rate, process deviation rate, and environmental deviation rate. The first-order pose deviation is represented using a state vector, which is calculated using a Kalman filter algorithm. Analysis based on the initial deviation rate, process deviation rate, and environmental deviation rate yields the translational deviation along the x, y, and z axes, θ. x θ y θ z It consists of six degrees of freedom: the rotational angle deviation of the component around the x-axis, y-axis, and z-axis.

[0036] A pose deviation prediction model based on Kalman filtering is designed, and dynamic estimation of the comprehensive pose deviation of components is achieved by fusing multi-source deviation rate data.

[0037] State vector θ represents the actual comprehensive pose deviation of the component that needs to be estimated, consisting of six degrees of freedom. Here, Δx, Δy, and Δz are the translational deviations along the x, y, and z axes of the component in the coordinate system of the construction site control network, based on the coordinates of the intended installation position; x θ y θ z These represent the rotational angle deviations of the component around the x-axis, y-axis, and z-axis, respectively.

[0038] The method for constructing a pose deviation prediction model based on Kalman filtering involves calculating the initial deviation rate δ using 3D laser scanning results. i .

[0039] X is obtained based on point cloud data from the holographic perception module. k The guessed value, the guessed value and the initial deviation rate δ i Multiplying them yields the initial state vector X0. X0 and X k They are state vectors of the same dimension; construct the state prediction equation:

[0040] F is the state transition matrix. For simple models, F can be set as the identity matrix I; indicating that the deviation at the next time step is the same as that at the current time step, the covariance prediction equation is constructed as follows:

[0041] Where F is the state transition matrix, P k-1 Let F be the covariance matrix of the previous time step. T Q is the transpose of the state transition matrix. k Let Q be the process noise covariance matrix; the dynamic adjustment equation for the process noise covariance matrix is ​​Q. k =Q base +γ*|δ ek |*I; where Q base Basic process noise; δ ek The environmental deviation rate calculated at time k, γ is the scaling factor, which maps the environmental deviation rate obtained from vibration intensity to the noise level, I is the identity matrix, and Q is the environmental deviation rate calculated at time k. base γ is calibrated through offline simulation or historical data analysis to balance the system's response speed and stability; The observation vector equation is constructed as follows:

[0042] Where H is the observation matrix, which is set as the identity matrix I in this example, v k This is observation noise; in this example, R k =|δ pk |*I, where δ pk Let I be the process deviation rate calculated at time k, and R be the identity matrix. kRepresents the uncertainty of the measuring machine itself; calculate the Kalman gain:

[0043] ,

[0044] Finally, the state prediction equation is: .

[0045] Based on the calculation results of the Kalman filter algorithm, a state vector ΔX containing six degrees of freedom is obtained. k The vector contains information about a pose deviation. From the design specifications or process requirements, the maximum permissible deviation value corresponding to each degree of freedom of the state vector is read. Each degree of freedom of the state vector is divided by its corresponding permissible deviation value to obtain six dimensionless deviation ratios. The total deviation rate is the sum of the six deviation ratios.

[0046] Intelligent Decision Module: Based on the predicted component's single-stage pose deviation result, it constructs a deviation quantification model and generates pose correction decisions. The deviation quantification model is used to analyze and define the factors that have the greatest impact on the total deviation, approximating the total deviation rate as δ. i δ p δ e Linear combination: ΔX k ≈β i *δ i +β p *δ p +β e *δ e +ε, where β i ,β p ,β e Let δ be the regression coefficient to be solved. i δ p δ e These represent the initial deviation rate, process deviation rate, and environmental deviation rate, respectively, with ε being the residual term, representing the portion not explained by the model. Using the collected data, the coefficients β are fitted using the least squares method. i ,β p ,β e The magnitude of these coefficients directly reflects the average influence of the corresponding deviation rate on the total deviation, and the conditions for the generation of the factors with the greatest influence are adjusted first.

[0047] The deviation vector is analyzed in depth, transforming abstract numerical values ​​into explicit adjustment requirements. The system sensitivity model, which is either pre-stored internally or obtained through external interfaces, is invoked. This model is formed through detailed mechanical analysis or calibration using a large amount of experimental data, and records in detail the influence of each actuator's action on the component's pose. A set of actuator action commands is calculated so that when these actuators act together, the component's pose change can precisely offset the deviations of the six degrees of freedom, achieving accurate correction. This action command serves as the decision for this pose correction.

[0048] It is worth noting that the motion command settings must be combined with strict constraints based on the actual scenario, including stroke constraints, which require that the operating range of each actuator cannot exceed the physical limit; motion amplitude constraints, which require that the single adjustment amplitude cannot be too large to avoid sudden stress increase or loss of control of components; and coordination constraints, which require that some actions be executed in a specific order to ensure that the optimization results are executable in reality.

[0049] Execution and Control Module: Receives posture correction decisions, drives the execution device to complete the specific operations of the decisions, and uses built-in sensors to provide feedback on the completion status of the decision operations.

[0050] The system receives pose correction decisions via Ethernet and performs format verification on these decisions, checking for complete data fields, correct data types, and reasonable numerical ranges. If a format error is found, an error message is immediately sent, requesting a retransmission of the correct pose correction decision. If the verification passes, the system proceeds to the next step of the parsing process. Taking into account the physical limitations of the actuator, motion stability requirements, task efficiency requirements, and environmental obstacle information, the optimal motion path is determined. The planned motion path is smoothed to eliminate inflection points and abrupt changes, reducing changes in the actuator's speed and direction, minimizing impact on the mechanical structure, and improving motion accuracy and stability. Control commands are issued to the actuator in batches based on the length and complexity of the motion path. For short paths and simple movements, all commands can be issued at once; for long paths and complex movements, a batch issuance method is used, issuing a certain number of commands in each batch while simultaneously receiving feedback from the actuator in real time. This ensures that the next batch of commands is issued only after the previous batch has been completed, avoiding command backlog and execution chaos.

[0051] During the operation of the actuator, sensors installed on the actuator collect motion parameters of the actuator in real time and feed the collected data back in real time. The actual motion parameters of the actuator collected in real time are compared with the target motion parameters in the command to calculate the real-time motion error. Based on the magnitude and trend of the error, a PID control algorithm is used to generate an error correction command, which is sent to the actuator to adjust the motion state of the actuator in real time, so that the actual motion parameters continuously approach the target motion parameters and ensure execution accuracy.

[0052] If the execution result satisfies the condition that all degrees of freedom offsets are less than the predicted degrees of freedom offsets, the successful execution result is fed back to the decision system, and preparations are made to receive the next adjustment instruction. If the execution result does not meet the preset indicators, all information including abnormal execution time, execution location, execution fault code, and execution parameters is recorded in detail, and the execution failure result and information are fed back to the decision system. At the same time, according to the type of abnormal situation, the corresponding fault diagnosis process is initiated to analyze the cause of the fault and take corresponding repair measures. After the fault is eliminated, the adjustment instruction is re-executed.

[0053] Closed-loop verification and self-learning module: For installed components that have completed the decision-making operation, pose information is collected again, secondary pose deviations are identified, and the secondary pose deviation data is fed back for incremental learning to continuously optimize the pose deviation prediction algorithm and the pose deviation optimization decision algorithm.

[0054] After the execution and control module completes the adjustment operation, a 3D laser scanner is used to perform a full-range scan of the component and the proposed installation location, following the same sampling rules as the first acquisition, specifically the same number of sampling points, the same distribution of sampling points, and the same sampling frequency.

[0055] During the data acquisition process, the sensor maps the acquired pose data to a unified coordinate system in real time to obtain the pose data of the target object in the unified coordinate system. It checks for missing sampling points or abnormal data values. If any are found, the data acquisition for the corresponding feature dimensions is repeated until complete and valid secondary pose data is obtained. The deviation of the secondary pose data is calculated to obtain the secondary pose deviation. If the absolute value of the secondary pose deviation is less than the absolute value of the primary pose deviation, it indicates that the adjustment decision is effective and the pose deviation has been reduced. If the absolute value of the secondary pose deviation in some dimensions is greater than the absolute value of the primary pose deviation, this abnormal situation needs to be marked to provide key analysis data for subsequent model learning.

[0056] To determine the validity of the secondary pose deviation data: Based on the task accuracy requirements, set allowable thresholds for deviations in each dimension under a unified coordinate system. If the secondary pose deviations in all dimensions are within the allowable threshold range, the adjustment is deemed to meet the accuracy requirements. If any dimension deviation exceeds the allowable threshold, the adjustment is deemed to have failed to meet the standards, and the secondary pose deviation and pose correction decision should be treated as key samples for focused monitoring.

[0057] Each pose correction decision is transformed into a learning opportunity. Through incremental learning, pose deviations during operation are continuously reduced, ultimately achieving fully automated high-precision installation with near-zero error. The self-learning model is input with primary pose deviation, pose correction decision, and secondary pose deviation. Dimensional data where the secondary pose deviation is greater than the primary pose deviation are marked as abnormal data, and vice versa.

[0058] The self-learning model is trained using both normal and abnormal data. It is trained with a data chain consisting of the current state (secondary pose deviation), the decision instruction (pose correction decision), and the predicted deviation (primary pose deviation), making its predictions increasingly closer to the physical response. The system provides feedback on the reduction in deviation, with the magnitude of the feedback positively correlated with the difference between the absolute values ​​of the primary and secondary pose deviations. Negative feedback is given to pose correction decisions for abnormal data, and positive feedback is given to pose correction decisions for normal data. Through continuous iteration, given a primary pose deviation, the model selects the pose correction decision that yields the maximum positive feedback. The presence of abnormal data helps the model quickly eliminate ineffective or harmful strategies.

[0059] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other.

[0060] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A steel structure construction precision control system based on multi-data fusion, characterized in that, Comprise: Full state perception module: use integrated sensor network, comprehensive collection component construction precision data, specifically by three-dimensional laser scanner collected component initial processing size data, by GNSS receiver, tilt sensor and force sensor collected component real-time spatial displacement data, and by vibration sensor collected component installation environment influence data; The deviation calculation module calculates the initial deviation rate δ of the component installation process based on the component initial machining size data, the component real-time spatial displacement data and the component installation environment influence data i , the process deviation rate δ p and the environment deviation rate δ e ; Initial deviation rate δ i The calculation method is: Wherein La is the component feature size extracted from the component point cloud data, and Lf is the corresponding feature size extracted from the point cloud data of the installation position. Process deviation rate δ p The calculation method is the product of the deviation rate of the component motion resultant velocity, the attitude angle deviation rate and the overall force deviation rate. Environmental bias rate δ e The calculation method is that when the measured component vibration frequency f m is greater than the safety threshold f a , when f m is less than or equal to the safety threshold f a , δ e =0; Smart prediction module: input initial deviation rate δ i , process deviation rate δ p and environmental deviation rate δ e , the Kalman filter algorithm is used to predict the one-time pose deviation of the component and the total deviation rate, wherein the one-time pose deviation is a six-degree-of-freedom state vector representing the deviation of the component relative to the tentative installation position, and the state vector includes x-axis, y-axis, z-axis translation deviation and rotation angle deviation θx, θy, θz around the x-axis, y-axis and z-axis; The intelligent decision module: according to the one-time pose deviation result of the prediction component, a deviation quantification model is constructed, and a pose correction decision is generated, wherein the deviation quantification model is used to analyze the influence weight of each deviation rate on the total deviation, and the model form is ΔX k ≈β i *δ i +β p *δ p +β e *δ e +ε, wherein β i , β p , β e are regression coefficients obtained by least square fitting, δ i , δ p , δ e are initial deviation rate, process deviation rate and environmental deviation rate respectively, and ε is a residual term. Execution and control module: receive pose correction decision, drive execution device to complete specific operation of decision, and adjust decision operation completion through built-in sensor feedback; Closed loop verification and self-learning module: for the installed component after executing the decision operation, collect pose information again, identify the second pose deviation, and feedback the second pose deviation data for incremental learning to continuously optimize the pose deviation prediction algorithm and the pose deviation optimization decision algorithm.

2. The steel structure construction precision control system based on multi-data fusion according to claim 1, characterized in that: The total deviation rate is calculated based on the state vector, and the maximum allowed deviation value corresponding to each degree of freedom of the state vector is divided by the allowed deviation value corresponding to each degree of freedom of the state vector to obtain six dimensionless deviation ratios, and the total deviation rate is the sum of the six deviation ratios.

3. The steel structure construction precision control system based on multi-data fusion according to claim 1, characterized in that: The bias quantification model is used to analyze and define the factors that have the greatest impact on the total bias, and the total bias rate is approximately expressed as δ i , δ p , δ e , a linear combination of: ΔX k ≈β i *δ i +β p *δ p +β e *δ e +ε, wherein β i , β p , β e are regression coefficients to be solved, δ i , δ p , δ e are initial bias rate, process bias rate and environmental bias rate respectively, and ε is a residual term representing the part that the model fails to explain; using the collected data, the coefficients β i , β p , β e are fitted by least squares method.

4. The steel structure construction precision control system based on multi-data fusion according to claim 1, characterized in that: The adjustment decision is based on the length and complexity of the motion path, and the control instructions are issued in batches to the execution device, for short path, simple motion, all instructions can be issued at one time; for long path, complex motion, the way of issuing in batches is adopted, a certain number of instructions are issued in each batch, and the feedback information of the execution mechanism is received in real time to ensure that the next batch of instructions is issued after the previous batch of instructions is executed.

Images