Method for monitoring and early warning of synchronous lifting of giant miter gate based on digital twin

Abstract

The application provides a kind of giant miter gate synchronous jacking digital twin monitoring and early warning method, the method constructs the digital twin model including geometric model, initial configuration model, support conversion correction model and fault mode library;Collect each jacking point displacement data, hydraulic pressure data, door body posture data and mechanical top state data;Through weight fusion and consistency check, the real-time state of the door body is obtained;When the door body is converted from door pivot support to synchronous jacking support, the jacking target configuration is formed;According to the configuration deviation, stress deviation and change trend, the risk level is determined;Screen candidate fault mode for early abnormality, perform fault injection and short-time evolution prediction, output graded warning and equipment handling action.The application can adapt to the configuration change caused by the support state conversion of the door body, identify jacking asynchronization, abnormal stress, abnormal foundation, hydraulic abnormality and mechanical top abnormality in advance, and improve the safety and controllability of jacking operation.

Description

Technical Field

[0001] This invention relates to the field of synchronous lifting of giant miter gates, and in particular to a digital twin monitoring and early warning method for synchronous lifting of giant miter gates. Background Technology

[0002] Giant miter gates are key hydraulic metal structures in navigation structures such as ship locks, serving long-term functions of water retention, navigation separation, and coordinated opening and closing. During maintenance of the bottom pivot, top pivot, sealing components, and lower structure of the gate, it is typically necessary to drain the water from the lock chamber and simultaneously lift the gate using jacking devices located at multiple stress points at the bottom of the gate. This gradually transfers the gate's weight from the pivot support state to the synchronous top support state, facilitating subsequent maintenance work. Due to the large weight, large span, limited bottom support area, and high center of gravity of giant miter gates, the force path, support state, and gate posture continuously change during the lifting process. Any asynchronous displacement, abnormal stress, abnormal foundation support, or failure of the mechanical support at any lifting point can lead to gate posture deviation, increased localized additional loads, structural jamming, or even instability.

[0003] Existing methods for lifting miter gates typically combine a main lifting strategy with synchronous zeroing and vertical alignment. Before lifting, workers obtain the gate's initial posture under pivot support by measuring the vertical lines from the front and sides of the gate, as well as the height difference or tilt angle. Then, a lifting device corresponding to one lifting point is selected as the main lifting device, and the posture of the remaining lifting points is corrected. The initial displacement readings of each lifting point are then uniformly zeroed, and synchronous lifting is carried out at a unified pace. This method can maintain the gate's posture stability to a certain extent during the lifting process, but after long-term operation, the gate often experiences sagging, lateral deviation, or local deformation, and its initial state is not ideally horizontal. If the ideal horizontal state is used directly as the control benchmark during lifting, it can easily create additional correction loads at each lifting point. If the initial posture under pivot support is always used as the fixed benchmark throughout the process, it is difficult to adapt to the structural changes caused by changes in force boundary conditions when the gate transitions from pivot support to synchronous lifting support.

[0004] Vertical line measurement can directly reflect whether the gate's posture has shifted, serving as a crucial safety verification basis during the lifting process of a miter gate. However, vertical line measurement primarily reflects the results after the gate's posture has already changed, making it difficult to promptly reveal early risks such as changes in force distribution at the lifting points, uneven foundation support, localized subsidence, hydraulic leakage, lifting jamming, and abnormal mechanical support. Especially when some lifting points experience increased displacement with insufficient thrust, continuously increasing thrust with limited displacement, or sudden increases or decreases in thrust, the gate's vertical line deviation may not yet be obvious. Relying solely on manual observation of the vertical line and experience-based adjustments makes it difficult to make early judgments about the source, development trend, and handling methods of risks.

[0005] Existing technology CN118815787B discloses a method and device for fault early warning of hydraulic systems based on digital twins. It establishes a digital twin model through the physical structure and historical operating data of the hydraulic system, collects monitoring data such as pressure, flow rate, temperature, vibration, and actuator position, and identifies anomalies and provides fault early warnings based on the working cycle, coordination relationship, and monitoring fitting relationship. This existing technology mainly targets the identification of the operating status and fault early warning of general hydraulic systems. Its monitoring focus is on the internal parameters and working cycle of the hydraulic system, making it difficult to address the changes in the gate configuration of a giant miter gate when it transitions from a pivot support to a synchronous top support. It also lacks a synchronous lifting monitoring basis based on the initial gate configuration, support transition correction amount, and target lifting configuration.

[0006] Existing technology CN122172579A discloses a synchronous control method for an integrated intelligent jacking platform in well tower construction. It employs a combination of digital twins, multi-source information sensing, synchronous control strategies, hydraulic execution feedback, cloud-edge collaboration, and safety early warning to monitor the displacement, pressure, tilt angle, and strain of the jacking platform during well tower construction. It also improves the operational safety of the jacking platform through synchronous control and graded early warning. This existing technology is applied to well tower construction jacking platforms, focusing on the calculation of control parameters and the coordinated control of hydraulic cylinders. It does not specifically address the actual initial configuration of the giant miter gate under pivot support, the configuration correction during the support transition stage, the projection relationship between the gate's attitude and the jacking point displacement, the mapping relationship between the jacking point thrust and the support state, or the correlation between the mechanical jacking state and the gate's safe attitude.

[0007] Therefore, the existing technology still lacks a monitoring and early warning method suitable for the synchronous jacking process of giant herringbone gates. This method can, on the basis of maintaining mature on-site processes such as main jacking strategy, synchronous zeroing, vertical line verification and mechanical jacking, perform correlation analysis on displacement data, hydraulic pressure data, gate posture data and mechanical jacking status data of each jacking point, identify the configuration changes of the gate body during the transition from gate pivot support to synchronous jacking support, and identify risks such as displacement coordination anomalies, thrust distribution anomalies, foundation support anomalies, hydraulic anomalies and mechanical jacking anomalies before the vertical line deviates significantly. Summary of the Invention

[0008] The main objective of this invention is to provide a digital twin monitoring and early warning method for synchronous lifting of a giant herringbone gate. The technical problem solved by this application is: how to achieve gate state reconstruction, early risk identification, fault prediction and graded early warning during the synchronous lifting process of a giant herringbone gate, based on the configuration changes caused by the transformation of the gate support state and the coupling relationship between the displacement, thrust, attitude and mechanical support state of the lifting point.

[0009] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a digital twin monitoring and early warning method for synchronous lifting of a giant miter gate, the method comprising: S1. Construct a digital twin model, which includes a geometric model, an initial configuration model under the support state of the door pivot, a support transformation correction model, and a failure mode library. S2. Collect multi-source monitoring data during the synchronous jacking process. The multi-source monitoring data includes displacement data of each jacking point, hydraulic pressure data, gate posture data, and mechanical jacking status data. S3. Based on multi-source monitoring data, perform weight fusion and consistency verification to calculate the real-time status of the gate; S4. During the stage when the gate body is transformed from the gate pivot support to the synchronous top support, the correction amount is calculated based on the support transformation boundary conditions, and the initial configuration model is updated to the lifting target configuration using the correction amount. S5. Compare the real-time status of the gate with the target configuration for lifting to obtain the configuration deviation, force deviation and their changing trends, and determine the initial risk level accordingly; S6. When the initial risk level or trend indicates an early anomaly, candidate fault modes are selected from the fault mode library, faults are injected into the real-time status of the door in the digital twin model and short-term evolution prediction is performed, the initial risk level is corrected based on the prediction results, a graded early warning is generated and the equipment handling actions are output.

[0010] In the preferred embodiment, in step S1, a gate coordinate system and a reference coordinate system at the bottom of the gate chamber are established, and the outer contour of the gate, the design center of gravity of the gate, the position of the gate pivot reference point, the spatial coordinates of each lifting point, and the spatial position of the attitude observation point are entered into the geometric model. An initial configuration model is established based on at least one of the following: front vertical offset data, side vertical offset data, height difference data, and tilt angle data obtained when the gate is in the gate pivot support state. The initial configuration model characterizes the reference point height and attitude parameters in two directions of the gate under actual stress. The support conversion correction model is used to characterize the configuration change caused by the change in force boundary conditions when the gate body is converted from the gate pivot support to the synchronous top support; the fault mode library includes observable residual modes and injectable disturbance parameters corresponding to the observable residual modes.

[0011] In the preferred embodiment, before step S2 begins, a lifting device corresponding to a lifting point is selected as the main lifting device, and the attitude of the remaining lifting points is corrected according to the initial configuration model so that the door body is restored to the initial attitude of the door pivot support state. After completing the attitude alignment, the initial readings of the displacement sensors at each lifting point are uniformly reset to zero to establish a lifting zero-position reference. After the lifting zero-position reference is established, each lifting point is lifted synchronously at a uniform rhythm so that the displacement relationship between each lifting point meets the requirements for maintaining the gate configuration.

[0012] In the preferred embodiment, in step S2, the displacement data of each lifting point is collected by a displacement sensor that corresponds to the stroke of the lifting device; Hydraulic pressure data is collected by pressure transmitters installed in the high-pressure chambers of each hydraulic cylinder, and converted into thrust data for each lifting point by combining the effective pressure area of ​​the hydraulic cylinder, return oil pressure or back pressure, and efficiency correction parameters obtained through static load testing; the gate attitude data includes at least one of vertical offset data, height difference data, laser ranging data, and tilt angle data. The mechanical top-holding status data includes at least one of the following: displacement data of the locking nut or gear mechanism, clearance data, limit switch status, locking signal, and top-holding motor coding data; All types of monitoring data are collected and stored according to a unified timestamp.

[0013] In the preferred embodiment, in step S3, a projection relationship from displacement to attitude is established based on the spatial coordinates of each lifting point and the displacement data of each lifting point, and the displacement change of each lifting point is projected as the rise of the door reference point and the attitude parameters in two directions. Based on the thrust data of each lifting point, a mapping relationship from the force to the support state is established, and the load sharing ratio of the lifting system bearing the weight of the gate body, the thrust distribution relationship of each lifting point, the thrust discrete state, the single-point unload state, the single-point overload state, and the local equivalent stiffness are calculated. The rise of the gate reference point, attitude parameters, load sharing ratio, thrust distribution relationship, and mechanical top protection status are taken as the real-time status of the gate.

[0014] In the preferred embodiment, in step S3, fusion weights are assigned to displacement data, thrust data, gate attitude data, and mechanical support status data according to sensor accuracy, installation reliability, data acquisition timeliness, and current consistency residual. The jacking point displacement data is used as high-frequency motion observation data, vertical offset data or tilt angle data is used as attitude reference observation data, and thrust data is used as support status and abnormal trend observation data to obtain the real-time status of the gate. When the deviation between the fusion result and any data source exceeds a preset consistency threshold, the corresponding data source is subjected to sensor verification, fusion weight reduction, or rejection. When hydraulic pressure data is abnormal and reliable thrust data cannot be obtained, the displacement data and gate posture data are retained for synchronous status monitoring, and the risk level is increased, the lifting speed is reduced, and a manual confirmation command is output.

[0015] In the preferred embodiment, in step S4, based on the thrust data at each lifting point and the weight of the gate body, according to the formula... Calculate the load sharing ratio of the support transfer ; when And continue At that time, it is determined that the door body has entered the support conversion stage; when , And continue At that time, confirm that the door body has completed the support conversion; According to the formula Calculate the average attitude value before entering the support transition phase, according to the formula. Calculate the average attitude value after the support transition phase is completed, according to the formula. Obtain the support conversion correction amount and follow the formula. Determine the target configuration for lifting; in, For the first The peak point is at The push of time, The number of climax points. For the weight of the door itself, To convert the load sharing threshold, To complete the load sharing threshold conversion, This marks the moment when the support transition phase begins. The moment when the support transition is completed.

[0016] In the preferred embodiment, in step S5, according to the formula Calculate the displacement synchronization deviation according to the formula. Calculate the attitude deviation between the real-time attitude of the gate and the configuration of the lifting target, according to the formula. Calculate the thrust dispersion coefficient at each lifting point, and based on... , and The displacement deviation growth rate, thrust mutation rate, and attitude drift rate were obtained respectively. When any lifting point experiences an increase in displacement without an increase in thrust or a decrease in local equivalent stiffness, it is determined that the lifting point has a risk of uneven foundation or local subsidence. When any lifting point exhibits restricted displacement and a continuous increase in thrust, it is determined that the lifting point is at risk of asynchrony or jamming.

[0017] In the preferred embodiment, in step S5, the initial risk level includes an alert level, a warning level, and a danger level; The prompt level corresponds to the equipment handling actions of automatically reducing the jacking speed, adjusting the non-main jacking follow-up amount, verifying the vertical line, and performing minor correction; the warning level corresponds to the equipment handling actions of pausing jacking, maintaining hydraulic pressure, correcting attitude, verifying the foundation support status, and activating mechanical jacking protection. The danger level corresponds to the equipment handling actions of stopping jacking, hydraulic pressure holding, mechanical jacking locking, and manual verification; when any jacking point experiences overload, unload, failure of mechanical jacking locking, rapid deterioration of attitude, or serious damage to data consistency, it is determined to be a danger level.

[0018] In the preferred embodiment, in step S6, according to the formula Construct observation residual feature vectors and compare them with residual feature templates in the fault mode library. Perform similarity calculations to filter candidate fault modes; For the jamming fault mode, according to the formula Perform fault injection; For hydraulic leakage failure modes, according to the formula Perform fault injection; For foundation unevenness or localized subsidence failure modes, according to the formula Implement fault injection, and make Growth occurs during linear forecasting; For sensor drift failure modes, according to the formula Perform fault injection; For each candidate failure mode, according to the formula Perform short-term evolution prediction and follow the formula. Obtain thrust variation; According to the formula Calculate the predicted risk index When the predicted risk index crosses the current risk level boundary, the initial risk level is increased and equipment handling actions matching the candidate failure modes are output.

[0019] This invention provides a digital twin monitoring and early warning method for synchronous jacking of a giant herringbone gate. This application establishes a digital twin model including a geometric model, an initial configuration model, a support transition correction model, and a fault mode library. The actual initial posture of the gate in the pivot support state is used as the modeling basis. When the gate transitions from pivot support to synchronous jacking support, the correction amount is calculated based on the support transition boundary conditions, updating the initial configuration model to the target jacking configuration. This avoids directly using the ideal horizontal state as the jacking control benchmark and also avoids mechanically using the initial configuration in the pivot support state as a fixed target throughout the entire process, thus making the posture judgment during the jacking process more closely match the actual stress state of the gate.

[0020] This application collects displacement data, hydraulic pressure data, gate posture data, and mechanical support status data at each jacking point. It then performs weighted fusion and consistency verification on the multi-source monitoring data. This ensures that changes in jacking point displacement reflect changes in gate posture, changes in hydraulic pressure reflect changes in force distribution and support status at the jacking points, and the mechanical support status reflects the backup load-bearing capacity of the jacking system. Therefore, it can identify potential risks such as asynchronous displacement, abnormal thrust distribution, uneven foundation, localized subsidence, hydraulic anomalies, and mechanical support anomalies even before significant vertical deviation occurs, improving the completeness and reliability of the monitoring results.

[0021] This application compares the real-time status of the gate with the target configuration for jacking to obtain configuration deviation, force deviation and their changing trends. According to the risk level, it outputs equipment handling actions such as reducing jacking speed, adjusting non-main jacking follow amount, pausing jacking, hydraulic pressure holding, attitude correction, mechanical jacking proximity protection, mechanical jacking locking and stopping jacking. This allows the early warning results to directly correspond to the executable operations at the jacking site, reducing the situation where there is only an alarm but no clear basis for handling.

[0022] This application filters candidate failure modes through a failure mode library and uses the real-time state of the gate as the initial state for failure injection and short-term evolution prediction in a digital twin model. It can perform advance simulations of situations such as asynchrony or jamming, abnormal force, uneven foundation or local subsidence, hydraulic abnormalities, mechanical jacking abnormalities, and sensor abnormalities. When the prediction results indicate that the risk may increase, it can correct the risk level in advance and output targeted handling actions, thereby enhancing the synchronous jacking process's ability to respond to early and sudden anomalies.

[0023] This application incorporates a combined safety mechanism of hydraulic jacking and mechanical jacking into the monitoring and early warning process. By monitoring the gap, position and locking status of the mechanical jacking mechanism, the mechanical jacking mechanism maintains a following state during normal jacking and enters a locked jacking state when the machine stops or a danger level warning is issued, providing continuous safety support for the synchronous jacking operation of the giant miter gate. Attached Figure Description

[0024] The present invention will be further described below with reference to the accompanying drawings and embodiments: Figure 1 This is a flowchart of the digital Li Sheng model establishment process of the present invention; Figure 2 This is a flowchart of the main top strategy and multi-source monitoring data acquisition process of this invention; Figure 3 This is a flowchart of the multi-source fusion and consistency verification process of this invention; Figure 4 This is a flowchart of the support conversion identification and lifting target configuration determination of the present invention; Figure 5This is a flowchart of the initial risk level determination and graded early warning and handling process of this invention; Figure 6 This is a flowchart of the fault mode screening, fault injection, and short-term evolution prediction process of the present invention. Figure 7 This is a schematic diagram of the synchronous lifting of the herringbone gate of the present invention. Detailed Implementation

[0025] Example 1 like Figure 1-7 As shown, a digital twin monitoring and early warning method for synchronous lifting of a giant miter gate is disclosed. The method includes: S1. Construct a digital twin model, which includes a geometric model, an initial configuration model under the support state of the door pivot, a support transformation correction model, and a failure mode library. S2. Collect multi-source monitoring data during the synchronous jacking process. The multi-source monitoring data includes displacement data of each jacking point, hydraulic pressure data, gate posture data, and mechanical jacking status data. S3. Based on multi-source monitoring data, perform weight fusion and consistency verification to calculate the real-time status of the gate; S4. During the stage when the gate body is transformed from the gate pivot support to the synchronous top support, the correction amount is calculated based on the support transformation boundary conditions, and the initial configuration model is updated to the lifting target configuration using the correction amount. S5. Compare the real-time status of the gate with the target configuration for lifting to obtain the configuration deviation, force deviation and their changing trends, and determine the initial risk level accordingly; S6. When the initial risk level or trend indicates an early anomaly, candidate fault modes are selected from the fault mode library, faults are injected into the real-time status of the door in the digital twin model and short-term evolution prediction is performed, the initial risk level is corrected based on the prediction results, a graded early warning is generated and the equipment handling actions are output.

[0026] The digital twin monitoring and early warning method for synchronous lifting of giant herringbone gates is based on the actual force changes of the gate body when it transitions from the gate pivot support state to the synchronous top support state.

[0027] Step S1 establishes a digital twin model. This digital twin model is not merely a 3D model displaying the gate's appearance, but a computational model used to continuously calculate the gate's attitude, support status, lifting synchronization relationship, and risk status. The geometric model records the gate's outline, design center of gravity, gate pivot reference point, spatial positions of the lifting points, and spatial positions of the attitude observation points. The initial configuration model records the gate's actual attitude under the gate pivot support state. The support transformation correction model is used to identify the configuration changes after the gate's weight is gradually transferred from the gate pivot to each lifting point.

[0028] The fault mode library is used to store the residual characteristics and injection parameters corresponding to asynchronous or jammed conditions, abnormal force, uneven foundation or local subsidence, hydraulic abnormalities, mechanical support abnormalities, and sensor abnormalities.

[0029] Step S2 collects displacement data, hydraulic pressure data, gate posture data, and mechanical support status data for each lifting point. Step S3 fuses the data from different sources to obtain the real-time status of the gate. The real-time status of the gate can be represented as follows: .

[0030] in, For a moment The real-time state vector of the gate. As the reference point for the gate body at time The increase in amount, These are the attitude parameters of the gate body around the first attitude direction. These are the attitude parameters of the gate body around the second attitude direction. For the lifting system at all times The load-sharing ratio that bears the weight of the door. This is the mechanical jacking state vector, where each element reflects the clearance, position, or locking state of the corresponding mechanical jacking mechanism. Step S4 updates the initial configuration in the door pivot support state to the jacking target configuration in the synchronous jacking support state using the support conversion correction amount.

[0031] Step S5 determines the initial risk level based on configuration deviation, stress deviation, and changing trends. Step S6 performs fault injection and short-term evolution prediction for early anomalies, so that the early warning results not only reflect the current anomaly but also the risk changes as the anomaly continues to develop.

[0032] In the preferred embodiment, in step S1, a gate coordinate system and a reference coordinate system at the bottom of the gate chamber are established, and the outer contour of the gate, the design center of gravity of the gate, the position of the gate pivot reference point, the spatial coordinates of each lifting point, and the spatial position of the attitude observation point are entered into the geometric model. An initial configuration model is established based on at least one of the following: front vertical offset data, side vertical offset data, height difference data, and tilt angle data obtained when the gate is in the gate pivot support state. The initial configuration model characterizes the reference point height and attitude parameters in two directions of the gate under actual stress. The support conversion correction model is used to characterize the configuration change caused by the change in force boundary conditions when the gate body is converted from the gate pivot support to the synchronous top support; the fault mode library includes observable residual modes and injectable disturbance parameters corresponding to the observable residual modes.

[0033] When establishing the geometric model, the gate coordinate system is denoted as... The reference coordinate system at the bottom of the gate chamber is denoted as .

[0034] The gate coordinate system describes the relative spatial relationships between the gate body, the gate pivot reference point, each lifting point, and the attitude observation points. The gate chamber bottom reference coordinate system describes the positional relationships between the lifting device, the gate chamber bottom plate measurement base point, and the gate body. The gate reference point can be the geometric center of the gate body, the projection point of the design center of gravity, or the corresponding position of the main top. The initial configuration model is based on the measured attitude of the gate body in the gate pivot support state, not on an ideal horizontal state as the initial target. The front vertical line and the side vertical line are used to obtain the attitude offset in two directions, respectively. Elevation difference data, laser ranging data, or tilt angle data are used to verify the vertical line measurement results.

[0035] According to the vertical attitude angle Obtained. Among them, This represents the attitude angle in the direction corresponding to the vertical line. This represents the horizontal offset of the vertical line on the measuring scale. This represents the effective height difference corresponding to the vertical measurement position. This relationship is used to convert the vertical offset into the gate's attitude angle under small angle conditions, enabling the initial configuration model to represent the stable configuration of the gate under actual stress conditions.

[0036] Each fault mode in the fault mode library includes an observable residual feature template and injectable perturbation parameters corresponding to that fault, enabling subsequent steps to select the appropriate prediction condition based on the current anomaly characteristics.

[0037] In the preferred embodiment, before step S2 begins, a lifting device corresponding to a lifting point is selected as the main lifting device, and the attitude of the remaining lifting points is corrected according to the initial configuration model so that the door body is restored to the initial attitude of the door pivot support state. After completing the attitude alignment, the initial readings of the displacement sensors at each lifting point are uniformly reset to zero to establish a lifting zero-position reference. After the lifting zero-position reference is established, each lifting point is lifted synchronously at a uniform rhythm so that the displacement relationship between each lifting point meets the requirements for maintaining the gate configuration.

[0038] The master jacking strategy is used to ensure a unified displacement and attitude reference during the synchronous jacking process. The master jacking can be determined based on the symmetry of the jacking point arrangement, the design center of gravity of the gate, or the location of the gate's reference point. Before jacking, the height of each jacking point is adjusted according to the initial configuration model and measurements of vertical lines, tilt angles, or height differences, restoring the gate to its actual initial attitude under the support of the gate pivot. After attitude alignment, the initial readings of the displacement sensors at each jacking point are uniformly zeroed, forming a jacking zero-position reference. The jacking zero-position reference is used to eliminate the influence of different initial sensor installation positions, stroke zero points, and the original drooping state of the gate. Subsequent synchronous jacking does not require all jacking points to have exactly the same absolute displacement, but rather requires that the displacement relationship of each jacking point meets the gate configuration maintenance requirements, allowing the gate to move upwards as a whole along its target jacking configuration, avoiding additional pitch, roll, or localized additional loads.

[0039] In the preferred embodiment, in step S2, the displacement data of each lifting point is collected by a displacement sensor that corresponds to the stroke of the lifting device; Hydraulic pressure data is collected by pressure transmitters installed in the high-pressure chambers of each hydraulic cylinder, and converted into thrust data for each lifting point by combining the effective pressure area of ​​the hydraulic cylinder, return oil pressure or back pressure, and efficiency correction parameters obtained through static load testing; the gate attitude data includes at least one of vertical offset data, height difference data, laser ranging data, and tilt angle data. The mechanical top-holding status data includes at least one of the following: displacement data of the locking nut or gear mechanism, clearance data, limit switch status, locking signal, and top-holding motor coding data; All types of monitoring data are collected and stored according to a unified timestamp.

[0040] Displacement data at each lifting point is collected by displacement sensors corresponding to the stroke of the hydraulic cylinder, reflecting the actual lifting height and relative synchronization relationship of each lifting point. Hydraulic pressure data is collected by a pressure transmitter installed in the high-pressure chamber of the hydraulic cylinder and converted into thrust data at the lifting point.

[0041] Thrust conversion relationship is .in, For the first The peak point at time The actual thrust, Number the jacking point. For the first Each lifting point corresponds to the high-pressure chamber of the hydraulic cylinder at a given time. Pressure This refers to the return oil pressure or back pressure of the hydraulic cylinder. For the first The effective pressure-bearing area of ​​each hydraulic cylinder For the first The efficiency correction factor for each hydraulic cylinder. The efficiency correction factor is determined through static load testing, pressure calibration, and the actual mechanical efficiency of the lifting device. When thrust needs to be expressed in tons of force, it is calculated according to... Perform the conversion. Among them, For the first The peak point at time The tonnage value, This is the conversion factor for gravitational acceleration.

[0042] The gate's attitude data includes at least one of the following: vertical offset data, elevation difference data, laser ranging data, and tilt angle data. The mechanical jacking status data includes the displacement data, clearance data, locking position signal, limit switch status, and jacking drive mechanism coding data of the jacking nut or gear mechanism.

[0043] All types of data are stored using a unified timestamp, so that displacement, thrust, attitude and mechanical support status at the same moment can correspond to the same gate working condition.

[0044] In the preferred embodiment, in step S3, a projection relationship from displacement to attitude is established based on the spatial coordinates of each lifting point and the displacement data of each lifting point, and the displacement change of each lifting point is projected as the rise of the door reference point and the attitude parameters in two directions. Based on the thrust data of each lifting point, a mapping relationship from the force to the support state is established, and the load sharing ratio of the lifting system bearing the weight of the gate body, the thrust distribution relationship of each lifting point, the thrust discrete state, the single-point unload state, the single-point overload state, and the local equivalent stiffness are calculated. The rise of the gate reference point, attitude parameters, load sharing ratio, thrust distribution relationship, and mechanical top protection status are taken as the real-time status of the gate.

[0045] The projection relationship between displacement and attitude is used to convert the displacement changes of multiple lifting points into the overall attitude change of the gate. To avoid using the same sign for planar coordinates and lifting displacements, in this embodiment, the planar coordinates of the lifting points are uniformly denoted as... and The displacement of the jacking point is denoted as .

[0046] The reference plane of the door body satisfies .in, For any position on the gate at time... height, and These are the planar coordinates of this position in the door coordinate system. and Here are the plane coordinates of the reference point for the door. This is the elevation of the door reference point. and These are the gate attitude parameters for two directions.

[0047] No. The displacement observation relationship of each apex point is as follows: .in, and For the first The planar coordinates of the apex point For the first The peak point at time The displacement measurement error. After writing the displacement relationship of each lifting point in matrix form, it can be expressed as: .in, The observation vector is composed of the displacements of each jacking point. The observation matrix is ​​established based on the spatial coordinates of each apex. Let be the vector of gate attitude parameters to be determined. This is the observation error vector.

[0048] The gate attitude parameters were obtained using the weighted least squares method, and the calculation relationship is as follows: .in, These are the estimated values ​​of the fused gate attitude parameters. The displacement observation weight matrix is... For the first The weights corresponding to the displacement data of each apex.

[0049] Lifting points with high displacement sensor accuracy, stable installation, and good consistency with other sensing channels have higher weights. When a displacement sensor is affected by loose installation, sudden changes, or external interference, its corresponding weight decreases. The mapping relationship between force and support state is used to obtain the load sharing ratio, thrust distribution relationship, thrust dispersion state, single-point unload state, single-point overload state, and local equivalent stiffness based on the thrust data of each lifting point, so that the real-time state of the gate simultaneously reflects the gate's attitude and support state.

[0050] In the preferred embodiment, in step S3, fusion weights are assigned to displacement data, thrust data, gate attitude data, and mechanical support status data according to sensor accuracy, installation reliability, data acquisition timeliness, and current consistency residual. The jacking point displacement data is used as high-frequency motion observation data, vertical offset data or tilt angle data is used as attitude reference observation data, and thrust data is used as support status and abnormal trend observation data to obtain the real-time status of the gate. When the deviation between the fusion result and any data source exceeds a preset consistency threshold, the corresponding data source is subjected to sensor verification, fusion weight reduction, or rejection. When hydraulic pressure data is abnormal and reliable thrust data cannot be obtained, the displacement data and gate posture data are retained for synchronous status monitoring, and the risk level is increased, the lifting speed is reduced, and a manual confirmation command is output.

[0051] Multi-source fusion employs displacement data, thrust data, attitude data, and mechanical support status data to jointly reconstruct the state. Displacement data, with its high update speed, is used to continuously update the rise of the gate reference point and attitude parameters. Vertical offset data or tilt angle data, with clear gravity reference significance, are used to correct the attitude parameters derived from displacement. Thrust data is used to identify the force distribution, support conversion process, and abnormal trends of the jacking system.

[0052] Mechanical jacking status data is used to identify whether the jacking mechanism maintains a normal following clearance and whether it completes locking under warning or shutdown conditions. After fusion, the deviation between the observed data and the estimated results is used as a consistency verification criterion. When the attitude change derived from displacement data is inconsistent with the vertical offset data and tilt angle data for a long period, it is determined that there may be displacement sensor drift, attitude sensor distortion, or loose sensor installation.

[0053] When tilt data is abnormal but vertical offset data is valid, the attitude reference is maintained using the vertical offset data. When vertical offset data is temporarily unavailable and displacement data is consistent with tilt data, both displacement and tilt data are used to maintain attitude monitoring.

[0054] When hydraulic pressure data is abnormal and reliable thrust data cannot be obtained, displacement data, attitude data, and mechanical jacking status data are retained for synchronous status monitoring. At the same time, the risk level is increased, the jacking speed is reduced, and manual confirmation is required to ensure that the failure of a single data source does not cause the entire monitoring process to fail.

[0055] In the preferred embodiment, in step S4, based on the thrust data at each lifting point and the weight of the gate body, according to the formula... Calculate the load sharing ratio of the support transfer ; when And continue At that time, it is determined that the door body has entered the support conversion stage; when , And continue At that time, confirm that the door body has completed the support conversion; According to the formula Calculate the average attitude value before entering the support transition phase, according to the formula. Calculate the average attitude value after the support transition phase is completed, according to the formula. Obtain the support conversion correction amount and follow the formula. Determine the target configuration for lifting; in, For the first The peak point is at The push of time, The number of climax points. For the weight of the door itself, To convert the load sharing threshold, To complete the load sharing threshold conversion, This marks the moment when the support transition phase begins. The moment when the support transition is completed.

[0056] The support transfer correction model identifies the process of transferring the gate's weight from the gate pivot to the lifting system through the load sharing ratio. The load sharing ratio is based on... calculate.

[0057] in, For a moment The load sharing ratio of the support transfer, This represents the total number of climax points. For the first The peak point at time The actual thrust, For the weight of the door itself. When the value gradually increases from a low level and then stabilizes, it indicates that the weight of the gate is now mainly borne by the synchronous lifting system. The criteria for entering the support transition stage are: And continue .in, For the support transition to enter the threshold, This refers to the duration of the support transition threshold. The condition for completing the support transition is... , And continue .

[0058] in, The threshold for completing the support conversion. The rate of change of the load sharing ratio over time. The load sharing ratio stability threshold, The duration for the support transition completion threshold. After entering the support transition phase, according to... Obtain the average posture value at the beginning of the transition.

[0059] After the support conversion is completed, according to Obtain the average pose value at the completion stage of the transition. Among them, For a moment The fused attitude parameters This marks the moment when the support transition phase begins. The moment when the support transition is completed, The length of the attitude sampling window at the beginning of the conversion phase. This is the length of the attitude sampling window during the transition completion phase.

[0060] Support conversion correction amount according to Obtained. Among them, This refers to the attitude correction amount generated during the transition from the gate pivot support state to the synchronous top support state. The lifting target configuration is based on... Confirmed. Among them, This serves as the target attitude for the subsequent stable jacking phase. Through the above processing, the subsequent jacking process uses the actual stable configuration under synchronous jacking support as a reference, avoiding directly using the initial configuration under the portal pivot support as the fixed target for the entire process.

[0061] In the preferred embodiment, in step S5, according to the formula Calculate the displacement synchronization deviation according to the formula. Calculate the attitude deviation between the real-time attitude of the gate and the configuration of the lifting target, according to the formula. Calculate the thrust dispersion coefficient at each lifting point, and based on... , and The displacement deviation growth rate, thrust mutation rate, and attitude drift rate were obtained respectively. When any lifting point experiences an increase in displacement without an increase in thrust or a decrease in local equivalent stiffness, it is determined that the lifting point has a risk of uneven foundation or local subsidence. When any lifting point exhibits restricted displacement and a continuous increase in thrust, it is determined that the lifting point is at risk of asynchrony or jamming.

[0062] Configurational deviations, force deviations, and their trends are used to identify early risks during synchronous jacking. Displacement synchronous deviations are calculated according to... Calculation. Among them, This represents the maximum relative displacement deviation among all jacking points at the current moment. For the first The current displacement of the apex point. This is the fused reference displacement formed by the displacements of each lifting point. Attitude deviations are calculated according to... calculate.

[0063] in, This refers to the deviation between the real-time attitude of the gate and the configuration of the lifting target. This represents the real-time posture of the merged gate. This is the configuration for the lifting target. The discrete thrust states at each lifting point are as follows: Calculation. Among them, For a moment The thrust dispersion coefficient, The standard deviation of the thrust at each lifting point. This represents the average thrust at each lifting point. An increase in the thrust dispersion coefficient indicates that the force distribution at each lifting point tends to be unbalanced. Single-point loss of load is calculated according to... judge.

[0064] in, The number of the jacking point to be determined. For the first The peak point at time The thrust, The threshold for the proportion of unloaded load. This represents the average thrust at each lifting point. Single-point overload is calculated according to... Judgment. Among them, The overload ratio threshold. This is the rated thrust of a single lifting device. The displacement deviation growth rate is calculated using... The thrust mutation rate is determined by... The attitude drift rate is determined using... judge.

[0065] in, The threshold for the displacement deviation growth rate. The thrust mutation rate threshold, This is the attitude drift rate threshold. When the displacement of the lifting point increases but the thrust does not increase, or when the local equivalent stiffness continues to decrease, it is determined that there is a risk of uneven foundation or local subsidence. When the displacement of the lifting point is limited but the thrust continues to increase, it is determined that there is a risk of asynchrony or jamming.

[0066] In the preferred embodiment, in step S5, the initial risk level includes an alert level, a warning level, and a danger level; The prompt level corresponds to the equipment handling actions of automatically reducing the jacking speed, adjusting the non-main jacking follow-up amount, verifying the vertical line, and performing minor correction; the warning level corresponds to the equipment handling actions of pausing jacking, maintaining hydraulic pressure, correcting attitude, verifying the foundation support status, and activating mechanical jacking protection. The danger level corresponds to the equipment handling actions of stopping jacking, hydraulic pressure holding, mechanical jacking locking, and manual verification; when any jacking point experiences overload, unload, failure of mechanical jacking locking, rapid deterioration of attitude, or serious damage to data consistency, it is determined to be a danger level.

[0067] The initial risk level includes at least a warning level, an alert level, and a hazard level. The warning level corresponds to a slight increase in displacement deviation, an emerging abnormal change in thrust, a tendency for attitude drift, or a decrease in sensor consistency, but not yet exceeding the control boundary. At the warning level, the control system reduces the jacking speed, adjusts the non-main jacking follow-up amount, performs minor correction, and prompts personnel to verify the vertical line.

[0068] The warning level corresponds to the working condition where the displacement synchronization deviation, attitude deviation, thrust dispersion state, or support conversion state has exceeded the preset control boundary.

[0069] Under the warning level, the control system suspends jacking, maintains hydraulic pressure, performs attitude correction, verifies the foundation support status, and controls the mechanical jacking mechanism to reduce the running follow-up gap.

[0070] The danger level corresponds to conditions such as single-point overload, single-point loss of load, rapid deterioration of the posture, failure of mechanical jacking lock, or severe disruption of multi-source data consistency. Under the danger level, the control system stops lifting, maintains hydraulic pressure, controls the mechanical jacking mechanism to eliminate gaps and lock, and simultaneously outputs a manual verification command. The mechanical jacking mechanism remains in a following state during normal lifting and enters a locked jacking state during shutdown or under the danger level, so as to assume the role of supporting the door body in the event of hydraulic system malfunction.

[0071] In the preferred embodiment, in step S6, according to the formula Construct observation residual feature vectors and compare them with residual feature templates in the fault mode library. Perform similarity calculations to filter candidate fault modes; For the jamming fault mode, according to the formula Perform fault injection; For hydraulic leakage failure modes, according to the formula Perform fault injection; For foundation unevenness or localized subsidence failure modes, according to the formula Implement fault injection, and make Growth occurs during linear forecasting; For sensor drift failure modes, according to the formula Perform fault injection; For each candidate failure mode, according to the formula Perform short-term evolution prediction and follow the formula. Obtain thrust variation; According to the formula Calculate the predicted risk index When the predicted risk index crosses the current risk level boundary, the initial risk level is increased and equipment handling actions matching the candidate failure modes are output.

[0072] The fault simulation uses the current real-time state of the gate as the initial state for prediction. First, the observation residual feature vector is constructed. .in, This is the current abnormal feature vector. This is a characteristic of displacement deviation. For thrust variation characteristics, These are characteristics of posture change. This represents the characteristic of local equivalent stiffness variation. Residual feature templates in the failure mode library Similarity calculation is performed. Among them, For the first Residual feature templates corresponding to different failure modes Number the fault mode. The jamming fault uses... Injection. Among them, For the first Predicted displacement of each fault jacking point The attenuation coefficient is the resistance coefficient. A value between 0 and 0.3 indicates a significant decrease or near-stop in the lifting speed. Hydraulic leakage faults are addressed using... Injection. Among them, For the first The pressure at the next predicted time point for each fault escalation point. The current resistance level at this peak point. To predict the time step, This is the pressure attenuation parameter. For foundation unevenness or localized subsidence, the following parameters are used. Injection. Among them, For the first The observed displacement of each apex point For the first The stroke displacement of a hydraulic cylinder For the first Each apex riser corresponds to a foundation subsidence amount, which increases according to a preset trend during the prediction process. Sensor drift faults are addressed using... Injection. Among them, For the first Each apex corresponds to the sensor's drift bias.

[0073] For candidate fault modes after fault injection, the jacking point displacement is determined according to... Perform short-term evolution prediction. Among them, For the first The displacement of the peak point at the next predicted time. For the first The displacement of each peak point at the current moment. For the first The escalation speed of the command at each escalation point To predict the time step, This represents the displacement change caused by the fault injection. Thrust changes are expressed as follows: An estimate is made. Among them, For the first Predicted thrust change at each apex rise point For the first Local equivalent stiffness at each apex. For the first The predicted displacement change at each apex. Local equivalent stiffness is obtained based on the relationship between thrust and displacement changes within the most recent time window. The predicted risk index is calculated according to... Calculation. Among them, To predict the risk index, , , and All are risk weights, and the sum of the four risk weights is one. This is the threshold for displacement synchronization deviation. For the current attitude deviation, This is the attitude deviation threshold. This represents the maximum thrust at each jacking point. The rated thrust of the lifting device, To predict the maximum attitude drift rate within the time window, This is the attitude drift rate threshold. When the predicted risk index crosses the current risk level boundary within a short prediction time window, the initial risk level is increased, and actions such as deceleration, correction, pressure maintenance, mechanical top locking, or shutdown that match the most likely failure mode are output.

[0074] Example 2 Further explanation in conjunction with Example 1, such as Figure 1-7 As shown in the figure, this embodiment provides a digital twin monitoring and early warning method for the synchronous lifting of a giant herringbone gate based on multi-source data fusion. Figure 1 This demonstrates the process of establishing a digital twin model. Figure 2 This demonstrates the main top strategy and the multi-source monitoring data acquisition process. Figure 3 This demonstrates the multi-source fusion and consistency verification process. Figure 4 This illustrates the process of support conversion identification and jacking target configuration determination. Figure 5 This demonstrates the initial risk level assessment and tiered early warning and response process. Figure 6 This demonstrates the process of fault mode screening, fault injection, and short-term evolution prediction. Figure 7 The diagram shows the on-site layout of the gate body, gate pivot support, hydraulic jacking device, mechanical jacking device, and various monitoring components during the synchronous jacking of the miter gate.

[0075] like Figure 7 As shown, multiple lifting points are set along the force-bearing positions of the gate body at the bottom. Each lifting point is equipped with a hydraulic lifting device and a mechanical jacking device. The hydraulic lifting device includes a hydraulic cylinder, a displacement sensor, and a pressure sensor. The displacement sensor is used to collect the lifting stroke of the corresponding lifting point, and the pressure sensor is set in the high-pressure chamber of the hydraulic cylinder to collect hydraulic pressure data. The mechanical jacking device is set correspondingly to the hydraulic lifting device. It is used to maintain a preset following gap during normal lifting and to eliminate the following gap and complete locking when the machine stops, pressure is maintained, or a danger level warning is issued. The mechanical jacking device is equipped with a gap sensor, a position sensor, a limit switch, and a locking position signal acquisition component to generate mechanical jacking status data.

[0076] A frontal vertical measurement system is installed on the front of the gate, while a side vertical measurement system, an inclinometer, and a laser rangefinder are installed on the sides. These systems together constitute the gate's attitude observation system. Reference points are set on the gate, located at its geometric center, the projected position of the design center of gravity, or the corresponding position on the main top. These reference points are used to represent the overall height increase and attitude changes of the gate. Figure 7 The door pivot support is used to support the weight of the door when it is in its initial state, and the hydraulic lifting device is used to gradually support the weight of the door during the synchronous lifting process.

[0077] like Figure 1 As shown, first establish the door coordinate system. and the reference coordinate system at the bottom of the gate chamber Door coordinate system Used to describe the relative spatial relationships between the gate's external outline, the gate pivot reference point, the lifting point, and the attitude observation point. Gate chamber bottom reference coordinate system. This is used to describe the relative positional relationships between the lifting device, the gate chamber floor, and the gate body. The outer contour of the gate body, the design center of gravity of the gate body, the position of the gate pivot reference point, the spatial coordinates of each lifting point, and the spatial position of each attitude observation point are entered into the geometric model.

[0078] When the gate is in the pivot-supported state, data on the front vertical offset, side vertical offset, elevation difference, and tilt angle are collected to establish an initial configuration model. The attitude angles corresponding to the vertical lines are... calculate. This represents the gate's attitude angle in the direction corresponding to the vertical line. This represents the horizontal offset caused by the vertical line relative to the measuring scale. This represents the effective height difference corresponding to the vertical measurement position. The actual posture of the gate in the pivot-supported state is obtained using the above method, without using the ideal horizontal state as the initial posture reference.

[0079] The support conversion correction model is used to characterize the configuration change that occurs when the gate body transitions from a pivot support to a synchronous top support. The fault mode library includes asynchronous or jamming fault modes, abnormal stress fault modes, uneven foundation or localized subsidence fault modes, hydraulic abnormality fault modes, mechanical top support abnormality fault modes, and sensor abnormality fault modes. Each fault mode is configured with a corresponding observation residual feature template and injectable disturbance parameters to facilitate subsequent fault simulation.

[0080] like Figure 2As shown, before the synchronous lifting begins, a hydraulic lifting device corresponding to one of the multiple lifting points is selected as the main lifting device. Based on the initial configuration model, the lifting devices corresponding to non-main lifting devices are calibrated to restore the gate to its initial posture under the support of the gate pivot. After the attitude calibration of each lifting point is completed, the initial readings of the displacement sensors at each lifting point are uniformly reset to zero to establish a zero-position reference for the lifting.

[0081] After establishing the zero-position benchmark, all lifting points are simultaneously lifted at a unified pace. Synchronous lifting does not mean maintaining the same absolute displacement at each lifting point, but rather ensuring a displacement relationship between the lifting points that matches the gate's configuration, allowing the gate to move upwards as a whole according to its actual configuration. During synchronous lifting, displacement data, hydraulic pressure data, gate posture data, and mechanical support status data for each lifting point are collected and stored using a unified timestamp.

[0082] The hydraulic pressure data at each lifting point is converted into thrust data, and the thrust conversion relationship is as follows: . For the first The peak point at time The thrust, For the first The high-pressure chamber of the hydraulic cylinder at each lifting point is at a certain time. hydraulic pressure, For the first The return oil pressure or back pressure of the hydraulic cylinder at each lifting point. For the first The effective pressure-bearing area of ​​each lifting point hydraulic cylinder For the first The efficiency correction factor for each lifting point hydraulic cylinder. The efficiency correction factor was obtained through static load test lifting and on-site calibration.

[0083] like Figure 3 As shown, the system receives jacking point displacement data, thrust data, gate attitude data, and mechanical jacking status data. A projection relationship from displacement to attitude is established based on the spatial coordinates of each jacking point. The peak point at time The displacement satisfies . For the first The peak point at time displacement, As the reference point for the gate body at time The increase in amount, and These are the attitude parameters of the gate along two attitude directions, respectively. and For the first The apex point in the gate coordinate system Planar coordinates in and The reference point of the door in the door coordinate system Planar coordinates in For the first The peak point at time Displacement measurement error.

[0084] The observation vector composed of the displacements of each apex point is denoted as... The real-time attitude parameter vector of the gate is denoted as The observation matrix established based on the spatial coordinates of the apex is denoted as... Then the projection relationship from displacement to attitude is expressed as: .in, Let be the displacement measurement error vector. The real-time attitude parameters of the gate are obtained using the weighted least squares method, and the calculation relationship is as follows: . These are the estimated values ​​of the real-time attitude parameters of the gate. The observation weight matrix is ​​determined based on sensor accuracy, installation reliability, data acquisition timeliness, and current consistency residuals.

[0085] The system establishes a mapping relationship between force and support state based on the thrust data at each lifting point, and calculates the load-sharing ratio of the gate's weight borne by the hydraulic lifting system. The load-sharing ratio is adopted using... calculate. For a moment The load sharing ratio, This represents the total number of climax points. The load-sharing ratio reflects the extent to which the weight of the door is gradually transferred from the door hinge support to the hydraulic lifting device.

[0086] The real-time status of the door is represented as follows . For a moment The real-time state vector of the gate. The mechanical jacking state vector is used to characterize the following gap, current position, and locking state of each mechanical jacking device.

[0087] After multi-source fusion, the consistency of the fusion results is verified. The observation residuals are used... calculate. For a moment The observed residual vector, This is the observation vector formed from the attitude observation data. The normalized consistency statistic uses... calculate. For a moment Consistency statistics To observe the noise covariance matrix. When When the preset consistency threshold is exceeded, the corresponding data source will undergo sensor verification, have its fusion weight reduced, or be removed.

[0088] When the tilt sensor data is abnormal but the plumb line data is valid, the plumb line data is used to maintain the gate's attitude reference. When the plumb line data is temporarily unavailable but the consistency between the displacement and tilt angle data meets the requirements, the displacement and tilt angle data are used to maintain gate attitude monitoring. When the hydraulic pressure data is abnormal and reliable thrust data cannot be obtained, the system retains the displacement data, gate attitude data, and mechanical jacking status data for synchronous status monitoring, and increases the risk level, reduces the jacking speed, and outputs a manual confirmation command.

[0089] like Figure 4 As shown, the system continuously inputs thrust data at each lifting point and real-time attitude parameters of the gate, and identifies the support conversion process based on the load sharing ratio. When the load sharing ratio reaches the support conversion entry threshold and remains so for a first preset duration, the gate is determined to have entered the stage of conversion from gate pivot support to synchronous top support. When the load sharing ratio reaches the support conversion completion threshold, the load sharing ratio change rate is less than the stability threshold, and remains so for a second preset duration, the gate is determined to have completed the support conversion.

[0090] At the beginning of the support transition phase, the fused attitude mean value within the first stable sampling window is acquired. During the support transition completion phase, the fused attitude mean value within the second stable sampling window is acquired. The support conversion correction amount is adopted. calculate. This refers to the attitude correction amount generated when the gate body changes from the gate pivot support state to the synchronous top support state. The lifting target configuration adopts... Sure. This serves as the target attitude configuration for the subsequent stable lifting phase. During the subsequent synchronous lifting process, the target lifting configuration is used as the evaluation benchmark for the real-time attitude of the gate, rather than the initial configuration under the gate pivot support state as the fixed benchmark for the entire process.

[0091] like Figure 5 As shown, the system compares the real-time state of the gate with the target lifting configuration, calculating the configuration deviation, force deviation, mechanical support state deviation, and their changing trends. Displacement synchronization deviation is calculated using... calculate. For a moment The maximum displacement synchronization deviation, For a moment The average displacement of each lifting point. The gate's attitude deviation is calculated using... calculate. For a moment The fused attitude parameters This represents the attitude deviation of the gate body relative to the lifting target configuration in real time.

[0092] Discrete thrust states at each lifting point are adopted calculate. For a moment The thrust dispersion coefficient, The standard deviation of the thrust at each lifting point. This represents the average thrust at each jacking point. An increase in the thrust dispersion coefficient indicates that the thrust distribution among the jacking points is becoming increasingly unbalanced.

[0093] When any lifting point experiences an increase in displacement without an increase in thrust, or a decrease in local equivalent stiffness, the system determines that the corresponding lifting point has a risk of foundation unevenness or local subsidence. When any lifting point experiences restricted displacement while thrust continues to increase, the system determines that the corresponding lifting point has a risk of asynchrony or jamming. The system determines the initial risk level based on configuration deviations, stress deviations, mechanical jacking status deviations, and their changing trends.

[0094] The initial risk levels include a warning level, an alert level, and a hazard level. The warning level corresponds to minor deviations or signs of abnormality; the system outputs actions such as reducing the jacking speed, adjusting the non-main jacking follow-up amount, performing minor corrections, and verifying the vertical line. The alert level corresponds to deviations exceeding limits or significant abnormalities; the system outputs actions such as pausing jacking, maintaining hydraulic pressure, correcting attitude, verifying the foundation support status, and controlling the mechanical jacking device to reduce the follow-up gap. The hazard level corresponds to severe abnormalities or rapidly deteriorating risks; the system outputs actions such as stopping jacking, maintaining hydraulic pressure, locking the mechanical jacking device, and performing manual verification.

[0095] like Figure 6 As shown, when the initial risk level or trend indicates early signs of anomaly, the system constructs an anomaly feature vector. . For abnormal feature vectors, This is a displacement anomaly characteristic. This is a thrust anomaly characteristic. These are characteristics of posture change. This represents the local equivalent stiffness variation characteristics. The system calculates the similarity between the abnormal feature vector and the residual feature template in the fault mode library to filter candidate fault modes.

[0096] For asynchronous or jamming fault modes, disturbance parameters indicating a decrease or cessation of displacement rate are injected into the corresponding jacking point. For hydraulic abnormal fault modes, disturbance parameters indicating pressure attenuation, reduced pressure holding capacity, or thrust fluctuation are injected into the corresponding jacking point. For foundation unevenness or localized subsidence fault modes, disturbance parameters indicating a continuous increase in foundation subsidence and a decrease in localized equivalent stiffness are injected into the corresponding jacking point. For mechanical jacking abnormal fault modes, disturbance parameters indicating lag in following, abnormal clearance, or locking failure are injected into the corresponding mechanical jacking device. For sensor abnormal fault modes, disturbance parameters indicating data drift, data jumps, or data loss are injected into the corresponding sensor channel.

[0097] In the short-term evolution prediction process, the first The predicted displacement of each apex at the next prediction time is adopted using calculate. For the first The predicted displacement of the peak point at the next prediction time. For the first The escalation speed of the command at each escalation point To predict the time step, The displacement change caused by the fault injection. The thrust change is predicted using the jacking point method. estimate. For the first Predicted thrust change at each apex rise point For the first Local equivalent stiffness at each apex. For the first The predicted displacement change at each apex.

[0098] The system calculates a predicted risk index for each candidate failure mode and identifies the candidate failure mode with the highest predicted risk index as the most likely failure mode. When the predicted risk index corresponding to the most likely failure mode crosses the current risk level boundary, the system adjusts the initial risk level to the corresponding predicted risk level and outputs equipment handling actions that match the most likely failure mode. Equipment handling actions include reducing the jacking speed, adjusting the non-main jacking follow-up amount, pausing jacking, hydraulic pressure holding, attitude correction, mechanical jacking locking, and stopping jacking.

[0099] Once the gate reaches the preset maintenance height, synchronous lifting stops. The hydraulic lifting device maintains pressure, while the mechanical jacking device eliminates the following gap and locks in place. The system reads the locking signals from each mechanical jacking device and performs a final safety check using plumb line measurement, tilt angle measurement, and real-time gate posture data. After confirming that the mechanical jacking status, hydraulic pressure holding status, and gate posture all meet safety requirements, the maintenance process proceeds to the gate hinge, waterstop, or lower gate structure.

[0100] The above embodiments are merely preferred technical solutions of the present invention and should not be considered as limitations on the present invention. The scope of protection of the present invention should be limited to the technical solutions described in the claims, including equivalent substitutions of the technical features described in the claims. That is, equivalent substitutions and improvements within this scope are also within the scope of protection of the present invention.

Claims

1. A digital twin monitoring and early warning method for synchronous lifting of a giant miter gate, characterized by: The method includes: S1. Construct a digital twin model, which includes a geometric model, an initial configuration model under the support state of the door pivot, a support transformation correction model, and a failure mode library. S2. Collect multi-source monitoring data during the synchronous jacking process. The multi-source monitoring data includes displacement data of each jacking point, hydraulic pressure data, gate posture data, and mechanical jacking status data. S3. Based on multi-source monitoring data, perform weight fusion and consistency verification to calculate the real-time status of the gate; S4. During the stage when the gate body is transformed from the gate pivot support to the synchronous top support, the correction amount is calculated based on the support transformation boundary conditions, and the initial configuration model is updated to the lifting target configuration using the correction amount. S5. Compare the real-time status of the gate with the target configuration for lifting to obtain the configuration deviation, force deviation and their changing trends, and determine the initial risk level accordingly; S6. When the initial risk level or trend indicates an early anomaly, candidate fault modes are selected from the fault mode library, faults are injected into the real-time status of the door in the digital twin model and short-term evolution prediction is performed, the initial risk level is corrected based on the prediction results, a graded early warning is generated and the equipment handling actions are output.

2. The digital twin monitoring and early warning method for synchronous lifting of a giant herringbone gate according to claim 1, characterized in that: In step S1, establish the gate coordinate system and the reference coordinate system at the bottom of the gate chamber, and input the gate outline, the gate design center of gravity, the position of the gate pivot reference point, the spatial coordinates of each lifting point and the spatial position of the attitude observation point into the geometric model. An initial configuration model is established based on at least one of the following: front vertical offset data, side vertical offset data, height difference data, and tilt angle data obtained when the gate is in the gate pivot support state. The initial configuration model characterizes the reference point height and attitude parameters in two directions of the gate under actual stress. The support conversion correction model is used to characterize the configuration change caused by the change in force boundary conditions when the gate body is converted from the gate pivot support to the synchronous top support; the fault mode library includes observable residual modes and injectable disturbance parameters corresponding to the observable residual modes.

3. The digital twin monitoring and early warning method for synchronous lifting of a giant herringbone gate according to claim 1, characterized in that: in Before starting step S2, select a lifting device corresponding to a lifting point as the main lifting device, and perform attitude correction on the remaining lifting points according to the initial configuration model to restore the door to its initial attitude under the door pivot support state. After completing the attitude alignment, the initial readings of the displacement sensors at each lifting point are uniformly reset to zero to establish a lifting zero-position reference. After the lifting zero-position reference is established, each lifting point is lifted synchronously at a uniform rhythm so that the displacement relationship between each lifting point meets the requirements for maintaining the gate configuration.

4. The digital twin monitoring and early warning method for synchronous lifting of a giant herringbone gate according to claim 1, characterized in that: in In step S2, the displacement data of each lifting point is collected by displacement sensors that are set up corresponding to the stroke of the lifting device; Hydraulic pressure data is collected by pressure transmitters installed in the high-pressure chambers of each hydraulic cylinder, and converted into thrust data for each lifting point by combining the effective pressure area of ​​the hydraulic cylinder, return oil pressure or back pressure, and efficiency correction parameters obtained through static load testing. The gate attitude data includes at least one of the following: vertical offset data, elevation difference data, laser ranging data, and tilt angle data; The mechanical top-holding status data includes at least one of the following: displacement data of the locking nut or gear mechanism, clearance data, limit switch status, locking signal, and top-holding motor coding data; All types of monitoring data are collected and stored according to a unified timestamp.

5. The digital twin monitoring and early warning method for synchronous lifting of a giant miter gate according to claim 4, characterized in that: in In step S3, a projection relationship from displacement to attitude is established based on the spatial coordinates of each lifting point and the displacement data of each lifting point, and the displacement change of each lifting point is projected into the lifting amount of the door reference point and the attitude parameters in two directions. Based on the thrust data of each lifting point, a mapping relationship from the force to the support state is established, and the load sharing ratio of the lifting system bearing the weight of the gate body, the thrust distribution relationship of each lifting point, the thrust discrete state, the single-point unload state, the single-point overload state, and the local equivalent stiffness are calculated. The rise of the gate reference point, attitude parameters, load sharing ratio, thrust distribution relationship, and mechanical top protection status are taken as the real-time status of the gate.

6. The digital twin monitoring and early warning method for synchronous lifting of a giant miter gate according to claim 5, characterized in that: in In step S3, the displacement data, thrust data, gate attitude data and mechanical support status data are assigned fusion weights according to sensor accuracy, installation reliability, data acquisition timeliness and current consistency residual. The jacking point displacement data is used as high-frequency motion observation data, the vertical offset data or tilt angle data is used as attitude reference observation data, and the thrust data is used as support status and abnormal trend observation data to obtain the real-time status of the gate. When the deviation between the fusion result and any data source exceeds the preset consistency threshold, the corresponding data source is subjected to sensor verification, fusion weight reduction or removal. When hydraulic pressure data is abnormal and reliable thrust data cannot be obtained, the displacement data and gate posture data are retained for synchronous status monitoring, and the risk level is increased, the lifting speed is reduced, and a manual confirmation command is output.

7. The digital twin monitoring and early warning method for synchronous lifting of a giant herringbone gate according to claim 1, characterized in that: in In step S4, based on the thrust data at each lifting point and the weight of the gate, the formula is used... Calculate the load sharing ratio of the support transfer ; when And continue At that time, it is determined that the door body has entered the support conversion stage; when , And continue At that time, confirm that the door body has completed the support conversion; According to the formula Calculate the average attitude value before entering the support transition phase, according to the formula. Calculate the average attitude value after the support transition phase is completed, according to the formula. Obtain the support conversion correction amount and follow the formula. Determine the target configuration for lifting; in, For the first The peak point is at The push of time, The number of climax points. For the weight of the door itself, To convert the load sharing threshold, To complete the load sharing threshold conversion, This marks the moment when the support transition phase begins. The moment when the support transition is completed.

8. The digital twin monitoring and early warning method for synchronous lifting of a giant miter gate according to claim 7, characterized in that: in In step S5, according to the formula Calculate the displacement synchronization deviation according to the formula. Calculate the attitude deviation between the real-time attitude of the gate and the configuration of the lifting target, according to the formula. Calculate the thrust dispersion coefficient at each lifting point, and based on... , and The displacement deviation growth rate, thrust mutation rate, and attitude drift rate were obtained respectively. When any lifting point experiences an increase in displacement without an increase in thrust or a decrease in local equivalent stiffness, it is determined that the lifting point has a risk of uneven foundation or local subsidence. When any lifting point exhibits restricted displacement and a continuous increase in thrust, it is determined that the lifting point is at risk of asynchrony or jamming.

9. The digital twin monitoring and early warning method for synchronous lifting of a giant miter gate according to claim 8, characterized in that: in In step S5, the initial risk level includes alert level, warning level, and danger level; The prompt level corresponds to the equipment handling actions of automatically reducing the jacking speed, adjusting the non-main jacking follow-up amount, verifying the vertical line, and performing minor correction; the warning level corresponds to the equipment handling actions of pausing jacking, maintaining hydraulic pressure, correcting attitude, verifying the foundation support status, and activating mechanical jacking protection. The danger level corresponds to the equipment handling actions of stopping jacking, hydraulic pressure holding, mechanical jacking locking, and manual verification; when any jacking point experiences overload, unload, failure of mechanical jacking locking, rapid deterioration of attitude, or serious damage to data consistency, it is determined to be a danger level.

10. The digital twin monitoring and early warning method for synchronous lifting of a giant miter gate according to claim 1, characterized in that: in In step S6, according to the formula Construct observation residual feature vectors and compare them with residual feature templates in the fault mode library. Perform similarity calculations to filter candidate fault modes; For the jamming fault mode, according to the formula Perform fault injection; For hydraulic leakage failure modes, according to the formula Perform fault injection; For foundation unevenness or localized subsidence failure modes, according to the formula Implement fault injection, and make Growth occurs during linear forecasting; For sensor drift failure modes, according to the formula Perform fault injection; For each candidate failure mode, according to the formula Perform short-term evolution prediction and follow the formula. Obtain thrust variation; According to the formula Calculate the predicted risk index When the predicted risk index crosses the current risk level boundary, the initial risk level is increased and equipment handling actions matching the candidate failure modes are output.

Images