A lattice column verticality control method based on multi-source fusion and prediction compensation

Abstract

The application provides a lattice column verticality control method based on multi-source fusion and prediction compensation, belongs to the verticality control technical field, and can realize real-time and accurate acquisition of the verticality data of the lattice column, and generate reliable three-dimensional deviation cloud diagrams by combining a novel anti-interference sensor, multi-point distributed installation, and advanced data transmission and processing technology; the intelligent hydraulic jack group control system based on real-time data feedback greatly shortens the adjustment time, avoids the over-limit problem of the insertion deviation caused by the adjustment lag, improves the verticality control precision of the lattice column, and ensures that the insertion deviation of the lattice column is controlled within a small range; the stability and reliability of the entire lattice column verticality control process are further improved, the uncertainty factors in the construction process are reduced, and the engineering quality and construction efficiency are improved by comprehensively optimizing the construction process, including drilling pretreatment, equipment debugging and calibration, lowering process monitoring, and concrete pouring process monitoring.

Description

Technical Field

[0001] This invention belongs to the field of verticality control technology, specifically relating to a method for controlling the verticality of lattice columns based on multi-source fusion and predictive compensation. Background Technology

[0002] Lattice columns are a common structural form in building construction, characterized by their high bending resistance. They are primarily constructed of reinforced concrete or steel sections, and feature high strength and a relatively small cross-section. Their cross-sections are typically designed to be biaxially or uniaxially symmetrical, improving bending efficiency by arranging materials away from the axis of inertia. They are widely used in factory frame columns, deep foundation pit support, and reverse construction methods. However, as mentioned in the prior art patent publication number "CN120367229A", verticality control of lattice columns is necessary during construction.

[0003] Current methods for controlling the verticality of lattice columns often rely on fixed inclinometers and their real-time data transmission to construct a three-dimensional deviation cloud map, which guides the hydraulic jack assembly in adjusting the verticality of the lattice column. However, in deep-hole or mud-slurry environments, sensors are susceptible to interference, leading to inaccurate or delayed data, making it impossible for the three-dimensional deviation cloud map to reflect the true state of the lattice column in real time. This lack of real-time performance often results in delayed adjustments by the hydraulic jack assembly, ultimately causing the lattice column insertion deviation to exceed limits (e.g., >20mm), severely impacting project quality and subsequent construction. Therefore, a new technical solution is urgently needed to address these issues. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method for controlling the verticality of lattice columns based on multi-source fusion and predictive compensation. This method overcomes the deficiencies of existing technologies in deep hole or mud environments, where sensor interference and data delay lead to insufficient real-time performance of the three-dimensional deviation cloud map, and the hydraulic jack assembly adjustment lags, resulting in excessive insertion deviation of the lattice column.

[0005] The present invention employs the following technical solution.

[0006] A method for controlling the verticality of lattice columns based on multi-source fusion and predictive compensation includes:

[0007] Step 1: Optimize the sensors for controlling the verticality of the lattice columns;

[0008] Step 2: Transmit and process the data collected by the sensors;

[0009] Step 3: Control the hydraulic jack assembly.

[0010] Furthermore, step 1 specifically includes: using an inclinometer based on fiber optic sensing for the sensors used to control the verticality of the lattice columns, which is connected to the central data processing system.

[0011] Furthermore, step 1 specifically includes: in the design of the sensor installation location, a multi-point distributed installation method is adopted, and multiple sensors are evenly set at different height positions of the lattice column.

[0012] Furthermore, step 2 specifically includes: sequentially connecting the central data processing system, Mesh network nodes, edge computing devices, and sensors for communication; deploying Mesh network nodes around the boreholes of the lattice column and at key locations set at the construction site of the lattice column; performing preliminary processing on the data collected by the sensors in real time on the edge computing devices; and then transmitting the pre-processed multi-sensor data to the central data processing system; and using an improved algorithm to fuse the pre-processed multi-sensor data to generate an accurate three-dimensional deviation cloud map.

[0013] Furthermore, in step 2, the central data processing system utilizes an improved algorithm to fuse the pre-processed multi-sensor data to generate an accurate three-dimensional deviation cloud map. This method specifically includes:

[0014] Step 2-1: Perform data fusion processing on the pre-processed data collected by multiple sensors;

[0015] Step 2-2: Generate a 3D cloud map based on the fused state values.

[0016] Furthermore, step 2-1 specifically includes:

[0017] Step 2-1-1: Perform data preprocessing on the data collected by the multi-sensor after preliminary processing;

[0018] Step 2-1-2: Perform noise suppression on the spatiotemporally aligned sensor data;

[0019] Step 2-1-3: Perform confidence weighting on each sensor.

[0020] Furthermore, step 2-1-1 specifically includes: performing spatiotemporal alignment on the pre-processed data collected by various sensors transmitted from the edge computing device through the Mesh network nodes, that is, using a cubic spline interpolation algorithm to correct the sampling time difference between different sensors, as shown in the following formula:

[0021] ;

[0022] in, The timestamp after synchronization, which is the corrected sensor number. Each sampling time, For the sensor Each original sampling time, For the maximum time deviation, This represents the number of sensors.

[0023] Furthermore, step 2-1-2 specifically includes:

[0024] An improved Kalman filter algorithm is used to filter noise from the spatiotemporally aligned sensor data. The core formula for noise filtering is:

[0025] State prediction equation: ;

[0026] Prediction error covariance: ;

[0027] Kalman gain: ;

[0028] State update equation: ;

[0029] Error covariance update: ;

[0030] in express The predicted state vector at time t. express The state transition matrix at time twilight is constructed based on the rigid body motion model of the lattice column. The time indicates the corrected sensor number. Each sampling time, express The control input matrix for the real-time correlation of the hydraulic jack adjustment amount. express The control input vector at time t, express The prediction error covariance matrix at time 1. express The process noise covariance matrix is ​​constantly corrected in real time by a mud density sensor. express Kalman gain at time step express The observation matrix that maps sensor data to state vectors at any given time. Indicates according to The observation noise covariance matrix of the sensor depth dynamically adjusted at any given time. express The vector of sensor observations at time 10:00. Represents the identity matrix.

[0031] Furthermore, step 2-1-3 specifically includes:

[0032] Constructing Bayesian network evaluation Confidence of each sensor , The calculation formula is:

[0033] ;

[0034] in, For the first Each sensor in environmental parameters The reliability probability density function is given below. For the number of sensors;

[0035] Then, the fused number is calculated using the following formula. Status values ​​of each sensor :

[0036] ;

[0037] in For the first One sensor value.

[0038] Furthermore, step 2-2 specifically includes:

[0039] The Kriging interpolation algorithm is used to generate a continuous 3D cloud map based on the fused state values. The deviation visualization formula is shown below:

[0040]

[0041] in, For spatial points The verticality deviation value, For the set number Interpolation weights for each sensor, For the first Status values ​​of each sensor , It is a Lagrange multiplier.

[0042] Furthermore, step 3 specifically includes:

[0043] Establish an intelligent hydraulic jack control system based on real-time data feedback. This means the hydraulic jacks are connected to a central data processing system. When the 3D cloud map shows a verticality deviation in the lattice column, the central data processing system automatically controls the hydraulic jacks to adjust them to the allowable range based on the direction and magnitude of the deviation.

[0044] The beneficial effects of the present invention are as follows, compared with the prior art:

[0045] This invention's novel anti-interference sensor, combined with multi-point distributed installation and advanced data transmission and processing technology, effectively solves the problems of interference and data delay in deep hole or mud environments. It can acquire the verticality data of the lattice column in real time and accurately, generating a reliable three-dimensional deviation cloud map. The intelligent hydraulic jack control system based on real-time data feedback can quickly and accurately adjust according to the real-time deviation of the lattice column, greatly shortening the adjustment time and avoiding excessive insertion deviation due to adjustment lag. This improves the verticality control accuracy of the lattice column, ensuring that the insertion deviation is controlled within a small range (e.g., ≤10mm). Through comprehensive optimization of the construction process, including borehole pretreatment, equipment debugging and calibration, monitoring of the lowering process, and monitoring of the concrete pouring process, the stability and reliability of the entire lattice column verticality control process are further improved, reducing uncertainties during construction and improving project quality and construction efficiency. Attached Figure Description

[0046] Figure 1 This is a flowchart of the lattice column verticality control method based on multi-source fusion and predictive compensation in this invention. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, other embodiments obtained by those skilled in the art without creative effort are all within the protection scope of this invention.

[0048] like Figure 1 As shown, a method for controlling the verticality of lattice columns based on multi-source fusion and predictive compensation includes:

[0049] Step 1: Optimize the sensors for controlling the verticality of the lattice columns;

[0050] In a preferred but non-limiting embodiment of the present invention, step 1 specifically includes: employing a novel anti-interference sensor. This sensor has a multi-layer shielding structure, which can effectively resist electromagnetic interference in the mud environment and complex stress interference in the deep hole environment. For example, the sensor used for controlling the verticality of the lattice column employs a fiber optic sensing-based inclinometer that is connected to a central data processing system. The fiber optic sensing-based inclinometer can be a Sitan TLX50-200D fiber optic gyroscope inclinometer, which uses optical signal transmission, is unaffected by electromagnetic interference, and has high accuracy and high stability. The central data processing system can be an industrial control computer.

[0051] In a preferred but non-limiting embodiment of the present invention, step 1 further includes: in the design of the sensor installation location, a multi-point distributed installation method is adopted, and multiple sensors are evenly set at different height positions of the lattice column, such as installing one sensor at the top, middle and bottom of the lattice column. In this way, real-time data of different parts of the lattice column can be obtained, thereby improving the accuracy and reliability of the data through data fusion algorithm.

[0052] Step 2: Transmit and process the data collected by the sensors;

[0053] In a preferred but non-limiting embodiment of the present invention, step 2 specifically includes: constructing an independent high-speed data transmission network using wireless mesh network technology. This technology features self-organizing networking and multi-hop transmission, enabling stable and high-speed data transmission in deep holes or mud environments. The central data processing system, mesh network nodes, edge computing devices, and sensors are sequentially connected for communication. Mesh network nodes can be deployed around the boreholes of the lattice column and at key locations set at the construction site of the lattice column to ensure that the data collected by the sensors can be quickly transmitted to the central data processing system. Regarding data processing, edge computing technology is introduced, offloading some data processing tasks to edge computing devices located close to the sensors. The edge computing devices can perform preliminary processing on the data collected by the sensors in real time, such as noise removal and outlier detection (noise removal and outlier detection can be handled using existing methods). Then, the pre-processed multi-sensor data is transmitted to the central data processing system. The central data processing system uses an improved algorithm to fuse the pre-processed multi-sensor data, quickly generating an accurate three-dimensional deviation cloud map.

[0054] In a preferred but non-limiting embodiment of the present invention, in step 2, the central data processing system uses an improved algorithm to fuse the pre-processed data collected by multiple sensors to quickly generate an accurate three-dimensional deviation cloud map. Specifically, this method includes:

[0055] The central data processing system employs an improved Kalman filter and Bayesian inference fusion algorithm to address noise interference, spatiotemporal synchronization deviation, and dynamic drift issues in multi-sensor data collected in deep-hole mud environments. The algorithm's innovation lies in three aspects: first, it introduces an environmentally adaptive noise covariance matrix dynamic correction mechanism to overcome the dependence of traditional Kalman filters on fixed noise models; second, it constructs a dynamic evaluation model of sensor confidence through a Bayesian network to achieve weighted fusion of heterogeneous sensor data; and third, it establishes a deviation prediction model based on the structural mechanical characteristics of the lattice column, generating deviation trend predictions 500ms in advance, thus allowing sufficient reaction time for hydraulic adjustments.

[0056] Step 2-1: Perform data fusion processing on the pre-processed data collected by multiple sensors;

[0057] In a preferred but non-limiting embodiment of the present invention, step 2-1 specifically includes:

[0058] Step 2-1-1: Perform data preprocessing on the data collected by the multi-sensor after preliminary processing;

[0059] In a preferred but non-limiting embodiment of the present invention, step 2-1-1 specifically includes: performing spatiotemporal alignment on the pre-processed data collected by each sensor transmitted from the edge computing device through Mesh network nodes, that is, using a cubic spline interpolation algorithm to correct the sampling time difference between different sensors, as shown in the following formula:

[0060] ;

[0061] in, The timestamp after synchronization, which is the corrected sensor number. Each sampling time, For the sensor Each original sampling time, The maximum time deviation represents the largest time difference between data collected by different sensors within the same sampling period. This represents the number of sensors.

[0062] Step 2-1-2: Perform noise suppression on the spatiotemporally aligned sensor data;

[0063] In a preferred but non-limiting embodiment of the present invention, step 2-1-2 specifically includes:

[0064] An improved Kalman filter algorithm is used to filter noise from the spatiotemporally aligned sensor data. The core formula for noise filtering is:

[0065] State prediction equation: ;

[0066] Prediction error covariance: ;

[0067] Kalman gain: ;

[0068] State update equation: ;

[0069] Error covariance update: ;

[0070] in express The predicted state vector at time (including the X, Y, and Z deviation values). express The state transition matrix at time twilight is constructed based on the rigid body motion model of the lattice column. The time indicates the corrected sensor number. Each sampling time, express The control input matrix for the real-time correlation of the hydraulic jack adjustment amount. express The control input vector at any given time (the control input vector includes the pressure change value of the hydraulic jack). express The prediction error covariance matrix at time 1. express The process noise covariance matrix is ​​constantly corrected in real time by a mud density sensor. express Kalman gain at time step express The observation matrix that maps sensor data to state vectors at any given time. Indicates according to Dynamic adjustment of sensor depth at any moment (as in deep hole environments) Increase the observation noise covariance matrix by 20%. express The vector of sensor observations at time 10:00. Represents the identity matrix.

[0071] Step 2-1-3: Perform confidence weighting on each sensor;

[0072] In a preferred but non-limiting embodiment of the present invention, step 2-1-3 specifically includes:

[0073] Constructing Bayesian network evaluation Confidence of each sensor , The calculation formula is:

[0074] ;

[0075] in, For the first Each sensor in environmental parameters The reliability probability density function under (mud viscosity, temperature, depth) is obtained through prior data obtained from field calibration. For the number of sensors;

[0076] Then, the fused number is calculated using the following formula. Status values ​​of each sensor :

[0077] ;

[0078] in For the first One sensor value.

[0079] Step 2-2: Generate a 3D cloud map based on the fused state values.

[0080] In a preferred but non-limiting embodiment of the present invention, step 2-2 specifically includes:

[0081] The Kriging interpolation algorithm is used to generate a continuous 3D cloud map based on the fused state values. The deviation visualization formula is shown below:

[0082]

[0083] in, For spatial points The verticality deviation value, For the set number Interpolation weights for each sensor, For the first Status values ​​of each sensor , To ensure unbiased estimation of the Lagrange multiplier, the cloud map color mapping uses the HSV color space. A vertical deviation value of 0-5mm is green, a vertical deviation value of 5-10mm is yellow, and a vertical deviation value >10mm is a red warning.

[0084] The technical effects of step 2 are shown below:

[0085] Environmental adaptability: through and The dynamic correction (correlated with real-time data from the mud hydrometer) enables the algorithm to maintain an angle measurement accuracy of 0.1° even in deep holes above 30m, which is 40% higher than the traditional algorithm.

[0086] Predictive fusion: By combining the elastic deformation model of the lattice column, a pre-deformation compensation term is introduced into the state equation to achieve early prediction of deviation trends and solve the problem of hydraulic adjustment lag.

[0087] Robust design: When a sensor fails, the Bayesian confidence model automatically reduces its weight to below 0.05, ensuring that the fusion results are not affected by a single point of failure, and extending the system's fault-free operation time to 1200 hours.

[0088] The parameter calibration method in step 2 is as follows:

[0089] The initial values ​​of the noise covariance matrix were obtained through ground-based no-load tests. Let it be diag[0.01,0.01,0.005]. Set according to sensor model (fiber optic inclinometer) =0.02).

[0090] Environmental parameters and The mapping relationship was established through orthogonal experiments, and the fitting formula is: , For depth (m), The viscosity of the mud (Pa·s) is where the lattice column is located.

[0091] Confidence weight The calculation is recalculated every 5 minutes to ensure adaptation to dynamic environmental changes.

[0092] Step 3: Control the hydraulic jack assembly.

[0093] In a preferred but non-limiting embodiment of the present invention, step 3 specifically includes:

[0094] A smart hydraulic jack control system based on real-time data feedback was established. This means the hydraulic jacks are connected to a central data processing system. When a 3D cloud map shows a verticality deviation in the lattice column, the central data processing system automatically controls the hydraulic jacks to adjust to the allowable range based on the direction and magnitude of the deviation. For example, if the lattice column tilts forward in the X direction, the central data processing system automatically controls the corresponding hydraulic jacks to increase pressure, pushing the lattice column backward. Simultaneously, it monitors the deviation changes in real time. When the deviation approaches the allowable range, it automatically reduces the adjustment force of the hydraulic jacks, achieving precise control. To ensure timely adjustment, the central data processing system has an early warning mechanism. When the deviation of the lattice column reaches a certain threshold (e.g., 10mm), an alarm is immediately issued, and the hydraulic jack assembly is activated for pre-adjustment to prevent further deviation.

[0095] In addition, the construction process for the lattice columns can be optimized: Before lowering the lattice columns, the borehole is pre-treated by using mud purification equipment to remove impurities and large particles, reducing interference from the mud on the sensors. Simultaneously, the sensors and hydraulic jacks are comprehensively debugged and calibrated to ensure optimal performance. During the lowering process, the data collected by the sensors and the lowering speed of the lattice columns are monitored in real time, and the lowering speed is adjusted according to the actual situation to avoid deviations caused by excessively rapid lowering and column swaying. When the column is lowered close to the design depth, the hydraulic jacks are used for fine-tuning to ensure the lattice columns accurately reach the design position and verticality requirements. During concrete pouring, the verticality changes of the lattice columns are continuously monitored, and if deviations occur due to the impact of concrete pouring, timely readjustment is made using the hydraulic jacks.

[0096] The beneficial effects of the present invention are as follows, compared with the prior art:

[0097] This invention's novel anti-interference sensor, combined with multi-point distributed installation and advanced data transmission and processing technology, effectively solves the problems of interference and data delay in deep hole or mud environments. It can acquire the verticality data of the lattice column in real time and accurately, generating a reliable three-dimensional deviation cloud map. The intelligent hydraulic jack control system based on real-time data feedback can quickly and accurately adjust according to the real-time deviation of the lattice column, greatly shortening the adjustment time and avoiding excessive insertion deviation due to adjustment lag. This improves the verticality control accuracy of the lattice column, ensuring that the insertion deviation is controlled within a small range (e.g., ≤10mm). Through comprehensive optimization of the construction process, including borehole pretreatment, equipment debugging and calibration, monitoring of the lowering process, and monitoring of the concrete pouring process, the stability and reliability of the entire lattice column verticality control process are further improved, reducing uncertainties during construction and improving project quality and construction efficiency.

[0098] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention without departing from the spirit and scope of the present invention. Any modifications or equivalent substitutions should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for controlling the verticality of lattice columns based on multi-source fusion and predictive compensation, characterized in that, include: Step 1: Optimize the sensors for controlling the verticality of the lattice columns; Step 2: Transmit and process the data collected by the sensors; Step 3: Control the hydraulic jack assembly; Step 1 specifically includes: using an inclinometer based on fiber optic sensing for the sensors used to control the verticality of the lattice columns, which is connected to the central data processing system; Step 1 also includes: in the design of the sensor installation location, a multi-point distributed installation method is adopted, and multiple sensors are evenly set at different height positions of the lattice column; Step 2 specifically includes: sequentially connecting the central data processing system, Mesh network nodes, edge computing devices, and sensors for communication; placing Mesh network nodes around the boreholes of the lattice column and at key locations set at the construction site of the lattice column; performing preliminary processing on the data collected by the sensors in real time on the edge computing devices; and then transmitting the pre-processed multi-sensor data to the central data processing system; using an improved algorithm, the central data processing system fuses the pre-processed multi-sensor data to generate an accurate three-dimensional deviation cloud map. In step 2, the central data processing system uses an improved algorithm to fuse the pre-processed multi-sensor data to generate an accurate 3D deviation cloud map. This method specifically includes: Step 2-1: Perform data fusion processing on the pre-processed data collected by multiple sensors; Step 2-2: Generate a 3D cloud map based on the fused state values; Step 2-2 specifically includes: The Kriging interpolation algorithm is used to generate a continuous 3D cloud map based on the fused state values. The deviation visualization formula is shown below: ； in, For spatial points The verticality deviation value, For the set number Interpolation weights for each sensor, For the first Status values ​​of each sensor , It is a Lagrange multiplier.

2. The method for controlling the verticality of lattice columns based on multi-source fusion and predictive compensation according to claim 1, characterized in that, Step 2-1 specifically includes: Step 2-1-1: Perform data preprocessing on the data collected by the multi-sensor after preliminary processing; Step 2-1-2: Perform noise suppression on the spatiotemporally aligned sensor data; Step 2-1-3: Perform confidence weighting on each sensor.

3. The method for controlling the verticality of lattice columns based on multi-source fusion and predictive compensation according to claim 2, characterized in that, Step 2-1-1 specifically includes: performing spatiotemporal alignment on the pre-processed data collected by various sensors transmitted from the edge computing device through the Mesh network nodes, that is, using a cubic spline interpolation algorithm to correct the sampling time difference between different sensors, as shown in the following formula: ； in, The timestamp after synchronization, which is the corrected sensor number. Each sampling time, For the sensor Each original sampling time, For the maximum time deviation, This represents the number of sensors.

4. The method for controlling the verticality of lattice columns based on multi-source fusion and predictive compensation according to claim 3, characterized in that, Step 2-1-2 specifically includes: An improved Kalman filter algorithm is used to filter noise from the spatiotemporally aligned sensor data. The core formula for noise filtering is: State prediction equation: ; Prediction error covariance: ; Kalman gain: ; State update equation: ; Error covariance update: ; in express The predicted state vector at time t. express The state transition matrix at time twilight is constructed based on the rigid body motion model of the lattice column. The time indicates the corrected sensor number. Each sampling time, express The control input matrix for the real-time correlation of the hydraulic jack adjustment amount. express The control input vector at time t, express The prediction error covariance matrix at time 1. express The process noise covariance matrix is ​​constantly corrected in real time by a mud density sensor. express Kalman gain at time step express The observation matrix that maps sensor data to state vectors at any given time. Indicates according to The observation noise covariance matrix of the sensor depth dynamically adjusted at any given time. express The vector of sensor observations at time 10:

00. Represents the identity matrix.

5. The method for controlling the verticality of lattice columns based on multi-source fusion and predictive compensation according to claim 4, characterized in that, Step 2-1-3 specifically includes: Constructing Bayesian network evaluation Confidence of each sensor , The calculation formula is: ； in, For the first Each sensor in environmental parameters The reliability probability density function is given below. For the number of sensors; Then, the fused number is calculated using the following formula. Status values ​​of each sensor : ； in For the first One sensor value.

6. The method for controlling the verticality of lattice columns based on multi-source fusion and predictive compensation according to claim 1, characterized in that, Step 3 specifically includes: A smart hydraulic jack control system based on real-time data feedback is established, which connects the hydraulic jacks to the central data processing system. When the 3D cloud map shows a verticality deviation in the lattice column, the central data processing system automatically controls the hydraulic jacks to adjust to the allowable deviation range according to the direction and magnitude of the deviation.

Images