Crane structure static and dynamic parameter monitoring method and system based on microwave sensing

Abstract

The application discloses a crane structure static and dynamic parameter monitoring method and system based on microwave sensing, and belongs to the technical field of crane deflection monitoring. The method comprises the following steps: step 1, microwave range finder installation and configuration; step 2, target retrieval and tracking; step 3, structure static parameter solving and benchmark establishment; step 4, deflection solving and reconstruction; step 5, quasi-static / vibration deflection separation and feature extraction; and step 6, evaluation, early warning and output. The application adopts a non-contact measurement mode based on microwave sensing, significantly improves the convenience and environmental adaptability of structure parameter monitoring, realizes integrated static and dynamic monitoring, effectively avoids low-frequency integral drift and zero-bias cumulative error compared with a displacement reconstruction method based on acceleration integration, significantly reduces construction complexity and safety risks, reduces long-term operation and maintenance hidden dangers such as sensor wear, loosening and protection failure, and thus reduces the whole life cycle maintenance cost of the system and improves the reliability.

Description

Technical Field

[0001] This invention belongs to the field of crane structure monitoring technology, specifically relating to a method and system for monitoring static and dynamic parameters of crane structures based on microwave sensing. Background Technology

[0002] Large cranes (lifting capacity ≥ 50 tons) are key equipment in ports, shipbuilding, bridge construction, and large-scale manufacturing, and their operational safety is highly correlated with economic benefits. During lifting, running, braking, and lowering operations, load-bearing components such as the main beam and cantilever not only bear the load weight, inertial force, and additional driving / braking loads, but also, in open-air and near-shore environments, are subject to the combined effects of wind loads, temperature gradients, salt spray corrosion, and damp heat. The random coupling of these multi-source loads easily induces significant deflection and dynamic vibration responses. Deflection, as a key indicator characterizing the overall stiffness and load-bearing capacity of a structure, can be used to identify abnormal behaviors such as overload and impact, and reveal potential problems such as loose connections, preload decay, local buckling, or decreased support stiffness, based on its absolute value, rate of change, baseline drift under the same conditions, and long-term growth trend. Therefore, constructing an online monitoring system capable of long-term stable operation in complex environments, achieving simultaneous acquisition, feature extraction, and risk warning of static and dynamic deflection, is of great significance for improving the operational safety of large cranes, optimizing operational efficiency, reducing maintenance costs, and supporting preventative maintenance and life-cycle management.

[0003] In existing technologies, the static parameter measurement of large crane structures mainly uses total stations and theodolites, but lacks dynamic deflection measurement capabilities. Dynamic deflection measurement / monitoring methods mainly include the following categories:

[0004] (1) Direct contact displacement measurement method

[0005] The method typically involves installing displacement gauges, LVDTs, or mechanical displacement meters at key structural locations. Using ground reference points or stable points as a reference, the structural displacement is converted into sensor output via mechanical connections or wires, thereby obtaining the displacement time history. The key to this method is establishing a stable geometric relationship between the measuring point and the reference point. However, in engineering implementation, it often requires additional supports and wiring, and is significantly affected by installation space, operational interference, wear and loosening, etc., and is prone to introducing additional errors under high-frequency dynamic conditions.

[0006] (2) Strain measurement and deflection inversion method

[0007] The method typically involves deploying resistance strain gauges or fiber optic strain sensors on the surface of the component to collect strain responses under different working conditions. A mechanical inversion model is then established based on beam bending theory, section parameters, and boundary conditions, and deflection is calculated from the strain. The key aspects of this method lie in the quality of sensor adhesion, temperature compensation, and the consistency between the inversion model and the actual structural boundaries. If there are adhesion failures, insufficient temperature drift compensation, or model errors, the reliability of the deflection inversion will be significantly affected.

[0008] (3) Optical / laser non-contact displacement measurement method

[0009] Methods typically employ laser ranging, total stations, or industrial camera-based visual target tracking to establish a line-of-sight measurement from the sensor to the target point, which is then converted into deflection. This usually relies on a clear and stable optical path and feature recognition conditions, requiring that the target be visible and the lighting controllable. However, in complex environments such as ports and steel mills, factors like rain, fog, dust, strong reflections, and obstructions often exist, leading to interrupted lines of sight, unstable recognition, or increased measurement noise, thus limiting monitoring stability.

[0010] (4) Vibration response measurement and displacement reconstruction method

[0011] The method typically uses accelerometers or IMUs to acquire the dynamic response of the structure, obtains the displacement through frequency domain integration or time domain integration, and then combines filtering and drift correction to achieve dynamic deflection estimation. The difficulty lies in the accumulation of low-frequency integral drift and zero bias error, making it difficult to reliably recover the static deflection or quasi-static displacement components; at the same time, it is quite sensitive to sensor installation coupling, noise suppression, and algorithm parameter selection.

[0012] The shortcomings of existing large crane deflection measurement / monitoring technologies are mainly reflected in the following aspects:

[0013] (1) Complex installation and deployment, high operation and maintenance costs: Existing monitoring systems generally require a lot of additional hardware and supporting construction. For example, contact displacement measurement requires the configuration of brackets, wires and fixed reference points, which has requirements for site space and construction conditions. In addition, there are problems such as wire wear, bracket loosening and sensor zero drift during long-term operation. Later maintenance is difficult, with a large workload and high cost.

[0014] (2) Insufficient environmental adaptability under complex field conditions: Optical / laser measurement is highly dependent on visibility conditions. Measurement accuracy drops significantly under conditions of backlight reflection, rain, fog, dust or obstruction of vision, and the target may even be lost. At the same time, in complex large metal structure scenarios, multipath reflection and strong clutter echo are easily generated, the target recognition is unstable and the measurement noise increases, which in turn affects the continuity and repeatability of deflection calculation and makes it difficult to meet the engineering requirements of long-term online monitoring.

[0015] (3) Inability to continuously measure static and dynamic deflection in an integrated manner: Existing solutions often present a fragmented state in terms of monitoring capabilities, namely "static is measurable but dynamic is insufficient" or "dynamic is measurable but static is distorted". Contact displacement sensors can directly measure static deflection, but due to the sensing method, they are not sensitive to dynamic response; acceleration-based dynamic monitoring requires displacement reconstruction through integration, but integral drift makes it difficult to reliably recover static deflection or quasi-static displacement components; visual measurement is difficult to stably capture dynamic displacement details and transient response under rapid conditions such as start-up, shutdown, and braking of hoisting, due to changes in frame rate, viewing angle and algorithm robustness, thus failing to form a unified measurement link for static and dynamic deflection covering the entire process.

[0016] (4) Lack of real-time assessment and early warning capabilities: Most existing solutions are still at the level of measurement value collection and display, lacking a closed-loop mechanism to transform deflection data into safety risk criteria. The system can usually only issue an alarm after the deflection has obviously exceeded the limit or the fault has become apparent, making it difficult to capture the initial signs of abnormality in a timely manner and achieve risk prevention.

[0017] In summary, existing crane deflection measurement technologies have significant limitations in terms of integrated static and dynamic measurement, stability in complex environments, engineering deployment and maintenance costs, and safety assessment and early warning capabilities. There is an urgent need to develop a new deflection monitoring method and system that can achieve continuous monitoring throughout the entire process, adapt to complex on-site environments, be easy to install and maintain, and have early warning capabilities. Summary of the Invention

[0018] The technical problem solved by this invention is to provide a method and system for monitoring the static and dynamic parameters of a crane structure based on microwave sensing. The aim is to achieve continuous monitoring, quantitative evaluation and timely early warning of the main structural parameters of the crane and the deflection response of key load-bearing components such as the main beam and boom under static and dynamic load conditions without contacting the crane structure or affecting its normal operation and safety protection.

[0019] Technical Solution: To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0020] A method for monitoring the static and dynamic parameters of a crane structure based on microwave sensing includes the following steps:

[0021] Step 1: Installation and configuration of the microwave rangefinder, specifically including:

[0022] Step 1.1: Fix the microwave rangefinder;

[0023] Step 1.2, Select and deploy measuring points: Select multiple target measuring points on the crane;

[0024] Step 1.3: Adjust the field of view of the microwave rangefinder: Adjust the pitch and azimuth angles of the microwave rangefinder so that the field of view of the microwave rangefinder covers all the ranges of all target measurement points reachable by the crane structure under all working conditions;

[0025] Step 2, target retrieval and tracking, specifically includes:

[0026] Step 2.1, Target Retrieval: Monitor the echo intensity within the field of view of the microwave rangefinder, and select signal points whose signal intensity reaches the threshold and matches the target measurement point;

[0027] Step 2.2, Target Tracking: The spatial resolution cell where the target measurement point is located and its neighborhood are used as the target tracking range. The target measurement point is continuously locked and a continuous measurement sequence is output.

[0028] Step 3: Structural static parameter calculation and benchmark establishment:

[0029] Step 3.1, Structural static parameter calculation: Extract the line-of-sight distance of all target measuring points in the spatial resolution unit, and combine the angle measurement information of the pitch angle and / or azimuth angle of the microwave rangefinder to calculate the camber, cantilever sag, span, base distance and trolley gauge parameters of the bridge crane, the verticality parameters of the tower crane, and the gauge, base distance, lifting range and amplitude of the gantry crane.

[0030] Step 3.2, Establish and calibrate the deflection baseline: Perform benchmark measurements under crane no-load and / or standard load conditions, and record the initial deflection parameters of each target measuring point;

[0031] Step 4, Deflection Calculation and Reconstruction, specifically includes:

[0032] Step 4.1 Deflection calculation: Extract the relative displacement increment of the line of sight direction of all target measurement points in the spatial resolution cell, and combine it with the elevation angle, azimuth angle and / or angle measurement information of the microwave rangefinder to project and map the relative displacement increment of the line of sight direction into the vertical deflection displacement time history of the gravity direction.

[0033] Step 4.2, Deflection Curve Reconstruction: Based on the vertical deflection displacement time history of multiple target measuring points, the deflection curve is reconstructed to obtain the dynamic deflection curve distribution of the crane structure during the monitoring period;

[0034] Step 5, Quasi-static / vibration deflection separation and feature extraction, specifically includes:

[0035] Step 5.1, Operating Condition Tagging and Event Segmentation: Integrate at least one of the following operating condition information from the crane: lifting capacity, lifting height, luffing angle, trolley position, running speed, braking status, and wind speed, and establish operating condition tags; and segment the vertical deflection displacement time history based on the operating condition tags and start / stop, acceleration / deceleration, or braking status.

[0036] Step 5.2, Quasi-static / Vibration Deflection Separation: Within the segment of the same working condition label, the quasi-static deflection component is obtained by using the quasi-static / vibration separation algorithm, and the vibration deflection component is obtained by subtracting the quasi-static deflection component from the vertical deflection displacement time history.

[0037] Step 5.3, Static / Dynamic Feature Parameter Extraction: Within the segment of the same working condition label, calculate the static stiffness of the crane based on the quasi-static deflection component and the crane span; calculate the camber of the main beam based on the main beam deflection curve reconstructed in Step 3.2; perform peak-to-peak value, root mean square (RMS), spectrum analysis, and frequency band energy calculation on the vibration deflection component, and calculate the dynamic amplification factor (DAF) based on the ratio of the peak vibration deflection to the quasi-static deflection reference value under the same working condition label, which is used to identify the dynamic amplification risk under start-stop and braking conditions;

[0038] Step 6: Evaluate the early warning and output, specifically including:

[0039] Step 6.1 Real-time threshold evaluation: Based on the preset quasi-static deflection limit, vibration deflection limit, DAF threshold, dominant frequency drift threshold and trend threshold rules, the real-time data is evaluated online.

[0040] Step 6.2, Long-baseline trend modeling under the same working condition: Under the same working condition label, perform long-term trend comparison modeling of the quasi-static deflection component and compare it with the baseline. Under the same working condition label, calculate the static stiffness index based on the lifting weight and quasi-static deflection component, and / or calculate the dynamic stiffness degradation index based on the main frequency drift and / or frequency band energy change.

[0041] Step 6.3, Multi-level early warning strategy: Set up three levels of early warning information output: prompt, warning, and alarm. Prompt, warning, or alarm is output based on the ratio of limit-type indicators to thresholds.

[0042] Preferably, the target measuring points are set at both ends, 1 / 4 span, mid-span, and 3 / 4 span of the main beam of the bridge crane; the target measuring points are set at the base of the tower body, the top of the tower body, the base of the boom, the middle of the boom, and the end of the boom of the tower crane; and the target measuring points are set at the end of the boom of the gantry crane, the lifting device, the center of the gantry leg, and the center of rotation. Targets are connected to the target measuring points. The targets are corner reflectors, spherical reflectors, metal strip targets, or magnetically detachable targets.

[0043] Preferably, in step 1.3, the field of view of the microwave rangefinder covers the entire range of all target measuring points reachable by the crane structure under all operating conditions.

[0044] Preferably, in step 2.1, under the background of strong metal reflection, multipath and clutter, a priori distance window constraint, constant false alarm rate CFAR threshold detection and peak clustering algorithm are used to lock the spatial resolution cell corresponding to the target or natural scattering point.

[0045] In step 2.2, Kalman filtering and / or α-β filtering are continuously performed on the spatial resolution cell where the target measurement point is located to achieve smooth updates and continuous locking of the target measurement point.

[0046] Preferably, in step 4.2, the deflection curve reconstruction adopts at least one of spline fitting, polynomial fitting, shape function fitting that satisfies the structural boundary conditions, finite element prior shape function parameterization fitting, finite element digital twin online correction, and machine learning regression model to obtain the deflection curve distribution of the target measurement point;

[0047] Preferably, in step 5.2, the adaptive filter cutoff frequency is engineered based on the structure's inherent frequency range and the operation cycle to ensure that the quasi-static deflection component reflects the quasi-static deformation and the vibration deflection component reflects the vibration response.

[0048] In step 5.2, the quasi-static / vibration separation algorithm employs at least one of bandpass filtering / moving average separation, wavelet decomposition, empirical mode decomposition (EMD), and variational mode decomposition (VMD) to obtain the quasi-static deflection component and the vibration deflection component.

[0049] Preferably, in step 5.2, based on obtaining the quasi-static deflection component, the deterministic deflection component caused by the crane mechanism motion is also eliminated to obtain the vibration residual; wherein, the deterministic deflection component is obtained by constructing a deterministic model from at least one of the working condition information of the amplitude angle, trolley position, lifting height, slewing angle, and running speed and its rate of change, and the deterministic model is at least one of regression model, state-space model, or parametric model based on mechanism kinematics / structural prior; the vibration residual is the remaining component after deducting the quasi-static deflection component and the deterministic deflection component from the vertical deflection displacement time history, and the vibration residual is used as the vibration deflection component for static / dynamic feature parameter extraction and dynamic amplification risk identification in step 5.3.

[0050] Preferably, in step 6.1, a duration criterion, a confidence criterion, and a multi-indicator consistency criterion are introduced to reduce the false alarm rate;

[0051] In step 6.1, the crane's lifting capacity, lifting height, amplitude angle, trolley position, running speed, braking status and wind speed are accessed to establish a working condition label and differentiated threshold system. Different limits and criteria are used under different working condition labels to avoid false alarms caused by fixed thresholds.

[0052] The online determination in step 6.1 also includes an anomaly identification algorithm. When the monitored value exceeds the limit or the anomaly continues to reach the preset condition, an alarm is triggered and the anomaly type is output. The anomaly types include quasi-static deflection exceeding the limit, dynamic amplification anomaly, main frequency drift anomaly, and suspected trend stiffness degradation.

[0053] Preferably, in step 6.3, for any limit-type indicator The limit-type index Quasi-static deflection Vibration deflection peak-to-peak value or root mean square Dynamic amplification factor , main frequency frequency drift and trend volume Define the ratio of at least one of them. , Limit value indicator In working condition label The corresponding threshold; when And if the confidence criterion is met, output "hint"; when And if the duration criterion and consistency criterion are satisfied, an "alert" is output; when or However, if the duration reaches the second time threshold, an "alarm" will be output.

[0054] The present invention also provides a monitoring system for static and dynamic parameters of a crane structure based on microwave sensing, comprising:

[0055] A microwave rangefinder is used to emit electromagnetic waves and receive echo signals reflected from a target measuring point, so as to output at least one of the following information related to the target measuring point: distance information, velocity information, angle information, and echo intensity information.

[0056] The installation support module is used to support and fix the microwave rangefinder, and to adjust the pitch and azimuth angles of the microwave rangefinder so that the microwave rangefinder beam covers the reachable space of the target measurement point under all working conditions.

[0057] A target reflection enhancement module is installed at the target measurement point to enhance the signal-to-noise ratio and / or distinguishability of the echo signal, thereby improving the quality of the measurement point data.

[0058] The working condition interface input is used to input at least one of the following working condition information: crane lifting capacity, lifting height, luffing angle, trolley position, running speed, braking status, and wind speed, and to generate working condition tags and start / stop / braking event segment information;

[0059] The target retrieval and tracking module is used to acquire the echo signal and perform signal processing to achieve target detection and tracking of the target measurement point, and output a continuous measurement sequence of the target measurement point. The target retrieval and tracking module performs target detection based on at least one of prior distance window constraint, CFAR threshold detection, peak clustering, or peak retrieval. The target retrieval and tracking module uses Kalman filtering and / or α-β filtering to track and update the continuous measurement sequence. The prior distance window and / or angle window are dynamically updated according to the working condition information to adapt to the spatial position changes of the target measurement point caused by structural movement.

[0060] The static parameter calculation module is used to calculate the camber, cantilever tilt, span, base distance and trolley gauge parameters of bridge and gantry cranes, as well as the verticality parameters of tower cranes, and the gauge, base distance, lifting range and amplitude of portal cranes, based on the line-of-sight distance of the spatial resolution unit where all target measurement points are located, and combined with the angle measurement information of the pitch angle and / or azimuth angle of the microwave rangefinder.

[0061] The deflection calculation and reconstruction module is used to extract the relative displacement increment of the target measuring point in the line of sight direction based on the continuous measurement sequence, and to project the relative displacement increment into the vertical deflection displacement time history in the gravity direction by combining the pitch angle, azimuth angle and / or angle measurement information of the microwave rangefinder; and to reconstruct the deflection curve distribution of the crane structure based on the vertical displacement time history of multiple target measuring points.

[0062] The quasi-static / vibration deflection separation and dynamic feature extraction module is used to obtain the quasi-static deflection component and static stiffness within segments of the same working condition label using a quasi-static / vibration separation algorithm, and to obtain the vibration deflection component by subtracting the quasi-static deflection component from the vertical deflection displacement time history; the quasi-static / vibration deflection separation and dynamic feature extraction module is also used to model and remove the deterministic deflection component caused by the motion of the crane mechanism to obtain the vibration residual, and to use the vibration residual as the vibration deflection component; the modeling of the deterministic deflection component is based on at least one of the working condition information of amplitude angle, trolley position, lifting height, slewing angle and running speed and its rate of change, and adopts at least one of regression model, state space model or parametric model based on mechanism kinematics / structural prior; and is used to calculate the peak-to-peak value, root mean square RMS, spectrum, frequency band energy and dynamic amplification factor (DAF) of the vibration deflection component;

[0063] The threshold assessment and early warning module is used to perform threshold assessment based on at least one of the quasi-static deflection component, vibration deflection component, dynamic amplification factor (DAF), dominant frequency drift, and trend index, so as to output warning information such as prompts, warnings, or alarms when deflection exceeds the limit, dynamic amplification is abnormal, or trend stiffness degradation is suspected. The threshold assessment and early warning module selects differentiated thresholds and / or performs real-time threshold assessment based on the working condition label. It is also used to perform long-baseline trend modeling under the same working condition label to identify potential risks of connection loosening, support stiffness changes, or structural stiffness degradation.

[0064] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0065] 1. Non-contact, all-weather, and stable monitoring capability: This invention adopts a non-contact measurement method based on microwave sensing, which is not sensitive to rain, fog, dust, low light, and a certain degree of obstruction. It can achieve long-term continuous operation in complex industrial environments such as ports, steel plants, and open-air construction, significantly improving the convenience and environmental adaptability of structural parameter monitoring.

[0066] 2. High-precision calculation and synchronous acquisition of static and dynamic deflection: This invention obtains quasi-static deflection and dynamic vibration displacement information of the structure simultaneously through a microwave rangefinder under the same measurement link, which can realize integrated static and dynamic monitoring. Compared with the displacement reconstruction method based on acceleration integral, this application effectively avoids low-frequency integral drift and zero bias accumulation error.

[0067] 3. Simplified system deployment and low maintenance cost: This invention does not require a large-scale deployment of contact displacement sensors and long-distance wiring in the structure, nor does it require complex mechanical reference devices. The displacement of the target measurement point can be completed by placing a microwave rangefinder and a passive target reflector in a suitable location, which significantly reduces construction complexity and safety risks. At the same time, it reduces long-term maintenance risks such as sensor wear, loosening and protection failure, thereby reducing the system's total life cycle maintenance cost and improving reliability.

[0068] 4. Real-time threshold assessment and hierarchical early warning capability: This invention can not only output the deflection time history of the target measurement point, but also further realize engineering indicators such as the maximum deflection at mid-span, deflection curve reconstruction, dynamic amplification factor, main frequency drift and trend indicators, which can be directly linked with structural limit criteria, working condition evaluation and operation and maintenance decisions; at the same time, it can construct a multi-level early warning strategy and support linkage output with the on-site monitoring platform or safety control system, thereby realizing risk prevention and safety closed-loop management. Attached Figure Description

[0069] Figure 1 This is a structural block diagram of a crane structure static and dynamic deflection monitoring system based on microwave sensing, according to an embodiment of the present invention.

[0070] Figure 2This is a schematic diagram of the microwave rangefinder setup in the embodiment;

[0071] Figure 3 This is a flowchart of an embodiment of a method for monitoring static and dynamic structural parameters of a crane structure based on microwave sensing;

[0072] Legend:

[0073] 1. Microwave rangefinder; 2. Microwave rangefinder line-of-sight azimuth adjustment device; 3. Tripod; 4. Ground base. Detailed Implementation

[0074] The present invention will be further illustrated below with reference to specific embodiments. These embodiments are implemented based on the technical solutions of the present invention, and it should be understood that these embodiments are only used to illustrate the present invention and are not intended to limit the scope of the present invention.

[0075] Example 1

[0076] A microwave sensing-based static and dynamic deflection monitoring system for crane structures, the system comprising:

[0077] The microwave rangefinder, using commercially available products, is used to emit electromagnetic waves and receive echo signals reflected from the target measuring point, so as to output at least one of the following information related to the target measuring point: distance information, velocity information, angle information, and echo intensity information. In this embodiment, the microwave rangefinder is a high-precision single-point or full-field microwave displacement and vibration measuring instrument from Shanghai Maiwei Technology Co., Ltd., with models such as HPSPMDS10-IM and MFVD50-I.

[0078] The mounting support module is used to support and fix the microwave rangefinder, and to adjust the pitch and azimuth angles of the microwave rangefinder, ensuring that the microwave rangefinder beam covers the reachable spatial range of the target measurement point under all operating conditions; such as Figure 2 As shown, in this embodiment, the installation support module uses a commercially available product. The installation support module includes a tripod 3 and a microwave rangefinder azimuth angle adjustment device 2. The microwave rangefinder azimuth angle adjustment device 2 uses an existing ball joint. The tripod 3 supports the microwave rangefinder 1 on the ground base 4.

[0079] The target reflection enhancement module, or target, is set at the target measurement point to enhance the signal-to-noise ratio and / or resolvability of the echo signal, thereby improving the quality of the measurement point data; the target can be an existing corner reflector, spherical reflector, metal strip target, or magnetically detachable target.

[0080] The working condition interface input is used to input at least one of the following working condition information: crane lifting capacity, lifting height, luffing angle, trolley position, running speed, braking status, and wind speed, and to generate working condition tags and start / stop / braking event segment information.

[0081] The target retrieval and tracking module is used to acquire echo signals and perform signal processing to achieve target detection and tracking of target measurement points, and output a continuous measurement sequence of target measurement points. The target retrieval and tracking module performs target detection based on at least one of prior distance window constraints, CFAR threshold detection, peak clustering, and peak retrieval. The target retrieval and tracking module uses Kalman filtering and / or α-β filtering to track and update the continuous measurement sequence, and the prior distance window and / or angle window are dynamically updated according to the working condition information to adapt to the spatial position changes of the target measurement points caused by structural movement.

[0082] The static parameter calculation module is used to calculate the camber, cantilever tilt, span, base distance, and trolley gauge parameters of bridge cranes, tower cranes, and gantry cranes, based on the line-of-sight distances of the spatial resolution units where all target measurement points are located, and combined with the angle measurement information of the pitch angle and / or azimuth angle of the microwave rangefinder.

[0083] The deflection calculation and reconstruction module is used to extract the relative displacement increment of the target measuring point in the line of sight direction based on the continuous measurement sequence, and to project the relative displacement increment into the vertical deflection displacement time history in the direction of gravity by combining the pitch angle, azimuth angle and / or angle measurement information of the microwave rangefinder; and to reconstruct the deflection curve distribution of the crane structure based on the vertical displacement time history of multiple target measuring points.

[0084] The quasi-static / vibration deflection separation and dynamic feature extraction module is used to obtain the quasi-static deflection component and static stiffness within segments of the same working condition label using a static-dynamic deflection separation method, and to obtain the vibration deflection component by subtracting the quasi-static deflection component from the vertical deflection displacement time history; the quasi-static / vibration deflection separation and dynamic feature extraction module is also used to model and remove the deterministic deflection component caused by the motion of the crane mechanism to obtain the vibration residual, and use the vibration residual as the vibration deflection component; the modeling of the deterministic deflection component is based on at least one of the working condition information of amplitude angle, trolley position, lifting height, slewing angle, running speed and its rate of change, and adopts at least one of the regression model, state-space model or parametric model based on mechanism kinematics / structural prior; and is used to calculate the peak-to-peak value, root mean square RMS, spectrum, frequency band energy and dynamic amplification factor DAF of the vibration deflection component.

[0085] The threshold assessment and early warning module is used to perform threshold assessment based on at least one of the following: quasi-static deflection component, vibration deflection component, DAF, dominant frequency drift, and trend index. When deflection exceeds the limit, dynamic amplification is abnormal, or trend stiffness degradation is suspected, at least one of the following outputs is a prompt, early warning, or alarm. The threshold assessment and early warning module selects differentiated thresholds based on the working condition label and / or performs real-time threshold assessment. It is also used to perform long-baseline trend modeling under the same working condition label to identify potential risks of connection loosening, support stiffness changes, or structural stiffness degradation.

[0086] This embodiment also provides a method for monitoring the static and dynamic parameters of a crane structure based on microwave sensing, including the following steps:

[0087] Step 1: Installation and configuration of the microwave rangefinder, specifically including:

[0088] Step 1.1: Fix the microwave rangefinder. Fix the microwave rangefinder to a rigid, stable and easy-to-maintain reference body (such as a ground base, track side column, factory column, or stable support of a non-measuring component of the crane) to establish a stable geometric relationship between the microwave rangefinder and the crane measuring point.

[0089] Step 1.2, Setting up and selecting measuring points: Select multiple target measuring points on the crane. For bridge cranes, set the two ends, 1 / 4 span, mid-span, and 3 / 4 span of the main beam as target measuring points. For tower cranes, set the base of the tower body, the top of the tower body, and the base, mid-span, and end of the boom as target measuring points. For gantry cranes, set the end of the boom, the lifting device, the center of the gantry legs, and the center of rotation as target measuring points. Targets are connected to the target measuring points. The targets are corner reflectors, spherical reflectors, metal strip targets, or magnetically detachable targets.

[0090] Step 1.3: Adjust the field of view of the microwave rangefinder: Adjust the pitch and azimuth angles of the microwave rangefinder so that the main lobe is aligned with the area where the target measurement point is located. The field of view of the microwave rangefinder covers all the ranges of all target measurement points that can be reached under all working conditions of the crane structure.

[0091] Step 2, target retrieval and tracking, specifically includes:

[0092] Step 2.1, Target Retrieval: Using prior distance window constraints, constant false alarm rate (CFAR) threshold detection, and peak clustering algorithm, spatial resolution cells corresponding to target points or natural scattering points are located in the range-angle information image of the echo. The spatial resolution cell is the smallest target area that the microwave rangefinder can resolve. When a target is set at the target measurement point, the spatial resolution cell corresponds to the target point. When no target is set at the target measurement point, the spatial resolution cell corresponds to a stable natural scattering point formed by the surface of the structure near the target measurement point.

[0093] In this embodiment, the prior distance window constraint is to determine the theoretical distance of each target measuring point based on the installation geometry of the microwave rangefinder, the theoretical position of the target measuring point, and / or the tracking result of the previous moment. An prior distance window (a pre-set range in which the target distance may exist) is set near the theoretical distance, and candidate target retrieval is performed within the prior distance window to suppress non-target echo interference from strong metal reflection, multipath, and clutter backgrounds.

[0094] In this embodiment, the constant false alarm rate (CFAR) threshold detection is performed by taking the sampling point of the distance to be detected as the detection unit within the prior distance window, selecting reference units on both sides of the sampling point and setting protection units, estimating the local clutter power based on the reference units and generating an adaptive detection threshold, and determining the detection unit as a candidate echo point when the echo intensity exceeds the threshold.

[0095] In this embodiment, the peak clustering algorithm performs peak clustering on candidate echo points along the distance axis, merging candidate points that are adjacent in distance or whose interval is less than a preset threshold into the same peak cluster, and using the position of the maximum peak or the energy weighting center of the peak cluster as the representative distance of the peak cluster; among multiple peak clusters, the peak cluster that is closest to the theoretical distance of the target measurement point and whose echo intensity meets the requirements is selected, and its corresponding spatial resolution unit is locked as the spatial resolution unit of the target measurement point.

[0096] Step 2.2, Target Tracking: The spatial resolution cell containing the target measurement point and its neighborhood are used as the target tracking range to achieve continuous locking of the target measurement point and output of the measurement sequence. During the tracking process, the distance and angle of the target measurement point are first predicted based on the estimated state of the previous time step to obtain the distance prediction value. and angle prediction value Then, the echo peaks are searched within the gated window, and the candidate peak with the smallest distance from the predicted value and whose echo intensity meets the threshold is selected as the observation at the current time. In order to suppress the observation fluctuations caused by noise, short-term occlusion or multipath, Kalman filtering and / or α-β filtering are continuously performed on the distance sequence and / or angle sequence of the spatial resolution cell where the target measurement point is located to achieve smooth update and continuous locking of the target measurement point. When the observation is missing within several consecutive sampling periods, the filtered predicted value is used for short-term maintenance and the search is expanded within the gated window until the target echo is recaptured or a re-search is triggered.

[0097] In this embodiment, the spatial resolution unit is jointly defined by the range resolution unit and the angular resolution unit, wherein the range resolution is... In the formula, Represents the speed of light. Indicates signal bandwidth; angular resolution is determined by the azimuth resolution of the microwave rangefinder. With pitch resolution Characterization, thus the size of each spatial resolution unit is approximately characterized as .

[0098] In this embodiment, the neighborhood is a gated window centered on the predicted target cell, and the gated window includes [the target cell] in the distance direction. Each range-resolved unit, in the angular direction including Each azimuth resolution unit and / or One pitch angle resolution unit; among which Choose 1 to 5, preferably 2 to 3. and / or Choose 1 to 3, preferably 1 to 2; the size of the gate window is preset according to the sampling period, the maximum possible displacement of the target measuring point under structural motion and vibration, and the measurement noise level.

[0099] In this embodiment, the Kalman filter uses a discrete state-space model to measure the distance. and / or azimuth Pitch angle For forecast updates, the key parameter is the process noise covariance. Covariance of measurement noise Process noise covariance Based on the uncertainties caused by structural vibration and operation; measure the noise covariance. Based on the ranging / angle measurement accuracy and echo signal-to-noise ratio settings of the microwave rangefinder; process noise covariance. and / or measure noise covariance It can be preset or adaptively adjusted according to the working condition label (select the corresponding parameter in the preset parameter table according to the working condition label). and Under steady-state conditions, a smaller To enhance smoothness, increase performance under dynamic conditions such as start-stop / braking. To enhance tracking performance; and to increase performance when the echo signal-to-noise ratio decreases or the risk of lock-off increases. (to reduce observation weight), thereby maintaining continuous locking of the target measurement point and suppressing measurement sequence jitter under clutter, multipath or short-term obstruction conditions.

[0100] In this embodiment, -β filtering with sampling period Extrapolate predictions and use the observed residuals according to gain. and Perform correction; The preferred value is 0.2 to 0.8. The preferred value is 0.05 to 0.5. and Based on the measurement noise level, echo signal-to-noise ratio, and the maximum possible speed and / or acceleration of the target measurement point, and based on the operating condition label, it switches between the start / stop / braking section and the steady-state section to improve smoothness and robustness while ensuring tracking performance.

[0101] Step 3: Structural static parameter calculation and benchmark establishment:

[0102] Step 3.1, Static Parameter Calculation: Extract the line-of-sight distance of all target measuring points within the spatial resolution unit. Combine this with the pitch and / or azimuth angle measurement information from the microwave rangefinder, and use trigonometric functions to calculate the camber, cantilever tilt, span, base distance, and trolley gauge parameters of the bridge crane, the verticality parameters of the tower crane, and the gauge, base distance, lifting range, and amplitude of the portal crane. (The principle and calculation method in this step are the same as the existing total station measurement method. By measuring the distance between the target measuring point and the line-of-sight direction of the total station, and combining the pitch and azimuth angle information, the above-mentioned static parameters of the crane can be calculated using trigonometric functions. This is a publicly available method and a conventional technical means for those skilled in the art.)

[0103] Step 3.2, Establish and calibrate the deflection baseline: Perform benchmark measurements under crane no-load and / or standard load conditions, and record the initial deflection parameters of each target measuring point;

[0104] Step 4, Deflection Calculation and Reconstruction, specifically includes:

[0105] Step 4.1 Deflection Calculation: Extract the relative displacement increment of the line of sight direction in the spatial resolution cell where all target measurement points are located, and combine it with the elevation angle, azimuth angle and / or angle measurement information of the microwave rangefinder to project and map the relative displacement increment of the line of sight direction into the vertical deflection displacement time history in the gravity direction.

[0106] In this embodiment, for each target measurement point The microwave rangefinder at the sampling time Output the distance measurement value of this measuring point. and the corresponding azimuth angle With pitch angle (Or output angle measurement information to determine the direction of the line of sight). Use the reference measurement time. distance For reference, the relative displacement increment in the line of sight direction is obtained as follows: Let the unit vector of the line of sight be denoted as The vertical unit vector is denoted as Let T denote the vector transpose. Then, the projection relationship between the displacement increment along the line of sight and the vertical deflection displacement increment is: Thus, the time history of the vertical deflection displacement of the target measuring point is obtained. When the angle information changes over time, and It can be updated at each sampling time to improve projection accuracy.

[0107] Step 4.2, Deflection Curve Reconstruction: In this embodiment, the coordinates of the measuring points are defined along the structural coordinate axes of the main beam or boom. At each sampling time Obtain a discrete sample set of deflection at multiple measurement points ( (The number of target measurement points), and reconstruct the continuous deflection curve based on this. In order to obtain the dynamic deflection curve distribution during the monitoring period.

[0108] The deflection curve reconstruction algorithm uses spline fitting:

[0109] For each sampling time ,by Construct cubic spline functions for nodes This ensures that at each node... It also ensures the continuity of the first and second derivatives; when the measurement noise is greater than the threshold (the measurement noise will vary with the working conditions and has no accurate range. In actual operation, the threshold corresponding to the working conditions is used to define the noise, and different thresholds are set for different working conditions), a smooth spline is used to add a curvature penalty term while minimizing the fitting error, so as to obtain a smooth and continuous deflection curve distribution.

[0110] Step 5, Quasi-static / vibration deflection separation and feature extraction, specifically includes:

[0111] Step 5.1, Operating Condition Labels and Event Segmentation: In this embodiment, at least one of the following operating condition information of the crane is accessed: lifting capacity, lifting height, luffing angle, trolley position, running speed, braking status, and wind speed, and an operating condition label is established; and the vertical deflection displacement time history is segmented based on the operating condition label and the start / stop, acceleration / deceleration, or braking status.

[0112] Step 5.2, Quasi-static / Vibration Deflection Separation: In this embodiment, the time history of the vertical deflection displacement of the target measuring point obtained in step 3 is denoted as... ,right Static deflection components are obtained by static-dynamic separation. With vibration deflection components and to Extract dynamic features and dynamic amplification factor (DAF).

[0113] In this embodiment, after obtaining the quasi-static deflection component Based on this, the deterministic deflection component caused by the movement of the crane mechanism is also analyzed. Remove them. To obtain vibration residual It is used as a vibration deflection component, a deterministic deflection component. The deterministic model is constructed from at least one of the following working condition information: amplitude angle, trolley position, lifting height, slewing angle, running speed and its rate of change. The deterministic model can be at least one of the following: regression model, state-space model or parametric model based on mechanism kinematics / structural priors.

[0114] The quasi-static / vibration separation algorithm uses bandpass filtering / moving average separation:

[0115] Time history of vertical deflection displacement of the target measuring point Static deflection components are obtained by performing digital filtering or moving average. The vibration deflection components are obtained in the form of residuals. Preferred: Static deflection component Vibration deflection components are obtained through low-pass filtering or moving average. Obtained through high-pass filtering, or by using For those focusing only on the vibration frequency band response, the time history of vertical deflection displacement can also be obtained. A bandpass filter is applied to obtain the vibration deflection component.

[0116] The engineering process for setting the filter cutoff frequency is as follows: Based on structural design data / finite element analysis or initial experiments, the natural frequency range of the structure (including at least the first dominant frequency) is identified. (and combine the typical cycles of lifting, luffing, and trolley movement operations), (Its characteristic frequency is approximately Determine the separation cutoff frequency , making The frequency is higher than the upper limit of the slowly varying frequency caused by the operation and lower than the lower limit of the main vibration frequency band, thus ensuring that the static deflection component reflects quasi-static deformation and the vibration deflection component reflects the vibration response. In engineering implementation, the spectral energy concentration of the separated vibration deflection component and the smoothness of the static deflection component can be checked and adjusted accordingly. Make minor adjustments.

[0117] Step 5.3: Static / Dynamic Feature Parameter Extraction

[0118] In this embodiment, under the rated lifting load condition, the trolley or lifting mechanism is placed in a preset measurement position, and the quasi-static deflection of the target measuring point at the mid-span of the main beam is extracted based on the quasi-static deflection component obtained in step 5.2. The sum of the lifting capacity and the self-weight of the trolley or hoisting mechanism is used as the equivalent static load. Calculate the static stiffness index of the main beam ; and quasi-static deflection With crane span Establish a deflection ratio criterion and determine the static stiffness compliance.

[0119] In this embodiment, based on the main beam deflection curve distribution obtained in step 4.2, the line connecting the support measuring points at both ends of the main beam is used as the baseline. The vertical geometric deviation of the target measuring point at the mid-span relative to the baseline is calculated and used as the camber of the main beam.

[0120] In this embodiment, the vibration deflection component Dynamic feature extraction is performed, including peak-to-peak value, root mean square (RMS) analysis, spectral analysis, and bandgap energy calculation. Spectral analysis can be performed using power spectrum estimation to obtain the dominant frequency and energy distribution. The dynamic amplification factor is calculated based on the ratio of the peak dynamic deflection to the reference static deflection value. , The reference value for static deflection is the value corresponding to the same working condition or event window (e.g., the mean value of static deflection components, the mean value of the steady-state segment, or the baseline value). DAF is used to identify the dynamic amplification risk under start-stop and braking conditions, and can be combined with the main frequency drift and frequency band energy anomaly for comprehensive judgment.

[0121] Step 6: Evaluate the early warning and output, specifically including:

[0122] Step 6.1, Real-time Threshold Assessment: In this embodiment, based on the quasi-static deflection component and vibration deflection component obtained in Step 5, the crane structural state is determined online. The process is as follows:

[0123] (1) Establishment of working condition labels and selection of differential thresholds: At least one of the following working condition information is collected: lifting capacity, lifting height, amplitude angle, trolley position, running speed, braking status and wind speed, to form working condition labels. (Divided into categories based on load, position / amplitude, and speed state (steady-state / acceleration / deceleration / braking)); each working condition is labeled. Pre-established or online updated threshold set ( , or , , ), and maintain the main frequency baseline. With quasi-static deflection baseline During online assessment, the current operating conditions should be determined first. Then, the corresponding threshold and baseline are selected for judgment.

[0124] (2) Sliding window index calculation: Calculate the quasi-static deflection within each sliding window. Vibration deflection peak-to-peak value or root mean square Dynamic amplification factor and clock speed And calculate the main frequency drift. With trend volume .

[0125] (3) Threshold comparison: based on the current working condition label Select the corresponding threshold for comparison; if or It is judged as excessive vibration deflection; if It is judged as an abnormal dynamic amplification; if It is judged to be an abnormal main frequency drift; if It was judged to be a suspected trend of stiffness degradation.

[0126] (4) Secondary confirmation and false alarm reduction: After secondary confirmation of the primary judgment result by introducing false alarm reduction criteria, the final judgment level is output. False alarm reduction criteria include at least one of the following:

[0127] Duration criterion: The exception must be met continuously. Only after confirming its effectiveness, =3s~5s; for short braking / start-stop events The time should be 1s to 3s.

[0128] Confidence criterion: When the echo signal-to-noise ratio or target lock quality is below a threshold (e.g.) When the lock loss rate exceeds the threshold, only record and downgrade the output, without directly triggering a high-level warning.

[0129] Multi-indicator consistency criteria: Only when "at least two indicators at the same measuring point are abnormal at the same time" or "the proportion of abnormalities among multiple measuring points of the same component reaches a threshold (e.g., ≥30%)" will it be upgraded to a warning or alarm.

[0130] Step 6.2, Long-baseline trend modeling under the same working condition: Under the same working condition label, perform long-term trend comparison modeling of the quasi-static deflection component and compare it with the baseline. Under the same working condition label, calculate the static stiffness index based on the lifting weight and quasi-static deflection component, and / or calculate the dynamic stiffness degradation index based on the main frequency drift and / or frequency band energy change.

[0131] In this embodiment, operating condition tags are used. Establish a long baseline library for the index, and save and update the quasi-static deflection baselines under this tag. Main frequency baseline and its statistical fluctuation range; baseline updates use exponential moving averages. , This represents the updated weights / smoothing coefficients of the exponential moving average. Preferably 0.01 to 0.1; relative deviation is used for comparison with the baseline. Alternatively, Z-score judgment can be used. When the deviation exceeds the trend threshold and meets the duration criterion, it is judged as a trend anomaly and "suspected trend stiffness degradation" is output.

[0132] In this embodiment, the static stiffness index can be calculated under the same working condition label. and contrast, For equivalent static load, The static stiffness baseline value; the dynamic stiffness degradation index is based on... Alternatively, the frequency band energy ratio can be obtained by comparing it with the baseline.

[0133] Step 6.3, Multi-level Early Warning Strategy

[0134] In this embodiment, three levels of output are set: prompt, warning, and alarm; for any limit-type indicator Quasi-static deflection Vibration deflection peak-to-peak value or root mean square Dynamic amplification factor , main frequency frequency drift and trend volume (at least one of them), define the ratio ;when And if the confidence criterion is met, output "hint"; when And if the duration criterion and the multi-indicator consistency criterion are satisfied, an "early warning" is output; when or However, if the duration reaches the second time threshold (the second time threshold is 10s to 30s), an "alarm" will be output.

[0135] The same tiered approach is used for frequency drift and trend indicators: a prompt is given when the threshold is reached by 0.8 times, a warning is given when the threshold is reached, and an alarm is triggered when the threshold is exceeded by 1.1 times or when there is a sustained significant (10s-30s) deviation.

[0136] Example 2

[0137] The difference from Example 1 is that the deflection curve reconstruction algorithm in step 4.2 of this example uses polynomial fitting:

[0138] For each sampling time Preset polynomial order (Preferred) To avoid overfitting (where N represents the number of target measurement points), construct... . For a moment polynomial coefficients, Represents coordinates, Represents the coefficient. Representing the power, using least squares to make Minimum, express exist The value at each position is used to obtain the coefficient vector. This allows us to obtain the deflection curve at that moment; for all Repeating this process will yield the dynamic deflection curve distribution.

[0139] Example 3

[0140] The difference from Example 1 is that the deflection curve reconstruction algorithm in step 4.2 of this example uses shape function fitting that satisfies the structural boundary conditions:

[0141] Select the set of shape functions that satisfy the structural boundary conditions. , The number of shape functions selected. For the first Each shape function is selected to naturally satisfy the structural boundary conditions, and a construction is performed. , For the first The shape function at time... The weighting coefficients for each sampling time. Solving for generalized coordinates using least squares or weighted least squares ( Depend on Composed of coefficient vectors of a row function expansion (also known as generalized coordinates), this minimizes the error at the measurement point, thus obtaining a deflection curve distribution that meets the boundary conditions.

[0142] Example 4

[0143] The difference from Example 1 is that the deflection curve reconstruction algorithm in step 4.2 of this example uses finite element prior shape function parameterization fitting:

[0144] A finite element model of the crane structure is pre-built, and prior shape functions related to the deflection shape (such as static deflection influence function, deflection basis shape or low-order modal shape under typical working conditions) are extracted to form the basis matrix. and build ; Generalized coordinates are given by the finite element model. These are the coefficients to be identified. For each sampling time... An observation equation is formed at the measurement point location. ( for exist (The value at the given location) is solved using least squares / weighted least squares. And thus in the entire region Output deflection curve distribution.

[0145] Example 5

[0146] The difference from Example 1 is that the deflection curve reconstruction algorithm in step 4.2 of this example uses online finite element digital twin correction:

[0147] Using a pre-established finite element model of the crane structure as a digital twin, the parameters to be corrected are defined. (e.g., equivalent stiffness, connection / support parameters, etc.) and / or Each sampling time Based on deflection observation at measuring points Construct a model-observation error, and apply recursive least squares, extended Kalman filtering, or unscented Kalman filtering to... and / or Online updates are performed to ensure the digital twin output matches the measured deflection; the updated digital twin model is then used to output the deflection curve distribution of the entire structure. And it is adjusted over time.

[0148] Example 6

[0149] The difference from Example 1 is that the deflection curve reconstruction algorithm in step 4.2 of this example uses a machine learning regression model:

[0150] Building a machine learning model for deflection field regression The input is a multi-point deflection vector. And / or operating condition information (lifting capacity, amplitude / trolley position, etc.), the output is a deflection vector on a discrete grid of structural coordinates. Or continuous function parameters; the machine learning model can take the form of Gaussian process regression, support vector regression, neural network, etc.; the training data comes from finite element simulation data and / or historical measured data; during online monitoring, the real-time measurement point deflection and working condition information are input into the machine learning model, and the dynamic deflection curve distribution during the monitoring period is directly output.

[0151] Example 7

[0152] The difference from Example 1 is that the quasi-static / vibration separation algorithm in step 5.2 of this example uses wavelet decomposition for separation:

[0153] Time history of vertical deflection displacement of the target measuring point Wavelet decomposition (Discrete Wavelet Transform (DWT) or Translation Invariant Wavelet Decomposition) is performed to obtain approximation coefficients and detail coefficients at different scales. The frequency bands corresponding to each scale are determined based on the sampling frequency and the number of decomposition levels, and the static deflection components are reconstructed from the low-frequency approximation. The vibration deflection components are obtained by reconstructing the high-frequency details. The number of decomposition layers and the static / dynamic boundary scale are determined based on the structure's natural frequency range and the operation cycle, so that the static deflection component covers the quasi-static low frequency band and the vibration deflection component covers the vibration response frequency band. When it is necessary to improve the transient fidelity of start-stop / braking, translation-invariant decomposition can be used and a higher resolution scale can be used within the event window.

[0154] Example 8

[0155] The difference from Example 1 is that the quasi-static / vibration separation algorithm in step 5.2 of this example uses empirical mode decomposition to separate EMD:

[0156] Time history of vertical deflection displacement of the target measuring point Empirical Mode Decomposition (EMD) was performed to obtain several intrinsic mode function (IMF) components and residual terms. Based on the dominant frequency (which can be characterized by instantaneous frequency or spectral peak value) and energy proportion of each IMF component, the static / dynamic boundary is determined: the high-frequency IMF components are superimposed and reconstructed into vibration deflection components. The low-frequency IMF components and / or residual terms Superimposed and reconstructed into static deflection components The boundary is determined based on the structure's natural frequency range and the operation cycle, so that the vibration deflection component mainly corresponds to the structure's vibration frequency band, and the static deflection component mainly corresponds to the quasi-static slow-changing part.

[0157] Example 9

[0158] The difference from Example 1 is that in step 5.2 of this example, the quasi-static / vibration separation algorithm uses variational mode decomposition (VMD) for separation:

[0159] Time history of vertical deflection displacement of the target measuring point Perform variational mode decomposition (VMD) to obtain Each narrowband modal component and its center frequency; based on the relationship between the center frequency of each mode and the natural frequency range of the structure, and the frequency corresponding to the operation cycle, the center frequency is set below the separation cutoff frequency. Modal superposition reconstructs static deflection components , with a center frequency higher than The modal superposition is reconstructed into vibration deflection components. ; The penalty parameters can be preset according to the frequency band coverage requirements or adjusted through energy concentration and reconstruction error verification to ensure that the static deflection component reflects the quasi-static deformation and the vibration deflection component reflects the vibration response.

[0160] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for monitoring the static and dynamic parameters of a crane structure based on microwave sensing, characterized in that, Includes the following steps: Step 1: Installation and configuration of the microwave rangefinder, specifically including: Step 1.1: Fix the microwave rangefinder; Step 1.2, Select and deploy measuring points: Select multiple target measuring points on the crane; Step 1.3: Adjust the field of view of the microwave rangefinder: Adjust the pitch and azimuth angles of the microwave rangefinder so that the field of view of the microwave rangefinder covers all the ranges of all target measurement points reachable by the crane structure under all working conditions; Step 2, target retrieval and tracking, specifically includes: Step 2.1, Target Retrieval: Monitor the echo intensity within the field of view of the microwave rangefinder, and select signal points whose signal intensity reaches the threshold and matches the target measurement point; Step 2.2, Target Tracking: The spatial resolution cell where the target measurement point is located and its neighborhood are used as the target tracking range. The target measurement point is continuously locked and a continuous measurement sequence is output. Step 3: Structural static parameter calculation and benchmark establishment: Step 3.1, Structural static parameter calculation: Extract the line-of-sight distance of all target measuring points in the spatial resolution unit, and combine the angle measurement information of the pitch angle and / or azimuth angle of the microwave rangefinder to calculate the camber, cantilever sag, span, base distance and trolley gauge parameters of the bridge crane, the verticality parameters of the tower crane, and the gauge, base distance, lifting range and amplitude of the gantry crane. Step 3.2, Establish and calibrate the deflection baseline: Perform benchmark measurements under crane no-load and / or standard load conditions, and record the initial deflection parameters of each target measuring point; Step 4, Deflection Calculation and Reconstruction, specifically includes: Step 4.1 Deflection calculation: Extract the relative displacement increment of the line of sight direction of all target measurement points in the spatial resolution cell, and combine it with the elevation angle, azimuth angle and / or angle measurement information of the microwave rangefinder to project and map the relative displacement increment of the line of sight direction into the vertical deflection displacement time history of the gravity direction. Step 4.2, Deflection Curve Reconstruction: Based on the vertical deflection displacement time history of multiple target measuring points, the deflection curve is reconstructed to obtain the dynamic deflection curve distribution of the crane structure during the monitoring period; Step 5, Quasi-static / vibration deflection separation and feature extraction, specifically includes: Step 5.1, Operating Condition Tagging and Event Segmentation: Integrate at least one of the following operating condition information from the crane: lifting capacity, lifting height, luffing angle, trolley position, running speed, braking status, and wind speed, and establish operating condition tags; and segment the vertical deflection displacement time history based on the operating condition tags and start / stop, acceleration / deceleration, or braking status. Step 5.2, Quasi-static / Vibration Deflection Separation: Within the segment of the same working condition label, the quasi-static deflection component is obtained by using the quasi-static / vibration separation algorithm, and the vibration deflection component is obtained by subtracting the quasi-static deflection component from the vertical deflection displacement time history. Step 5.3, Static / Dynamic Feature Parameter Extraction: Within the segment of the same working condition label, calculate the static stiffness of the crane based on the quasi-static deflection component and the crane span; calculate the camber of the main beam based on the main beam deflection curve reconstructed in Step 3.2; perform peak-to-peak value, root mean square (RMS), spectrum analysis, and frequency band energy calculation on the vibration deflection component, and calculate the dynamic amplification factor (DAF) based on the ratio of the peak vibration deflection to the quasi-static deflection reference value under the same working condition label, which is used to identify the dynamic amplification risk under start-stop and braking conditions; Step 6: Evaluate the early warning and output, specifically including: Step 6.1 Real-time threshold evaluation: Based on the preset quasi-static deflection limit, vibration deflection limit, DAF threshold, dominant frequency drift threshold and trend threshold rules, the real-time data is evaluated online. Step 6.2, Long-baseline trend modeling under the same working condition: Under the same working condition label, perform long-term trend comparison modeling of the quasi-static deflection component and compare it with the baseline. Under the same working condition label, calculate the static stiffness index based on the lifting weight and quasi-static deflection component, and / or calculate the dynamic stiffness degradation index based on the main frequency drift and / or frequency band energy change. Step 6.3, Multi-level early warning strategy: Set up three levels of early warning information output: prompt, warning, and alarm. Prompt, warning, or alarm is output based on the ratio of limit-type indicators to thresholds.

2. The method for monitoring static and dynamic parameters of a crane structure based on microwave sensing according to claim 1, characterized in that, In step 1.2, the two ends, 1 / 4 span, mid-span, and 3 / 4 span of the main beam of the bridge crane are set as target measuring points. The base of the tower body, the top of the tower body, and the base, mid-span, and end of the boom of the tower crane are set as target measuring points. The end of the boom of the gantry crane, the lifting device, the center of the gantry leg, and the slewing center are set as target measuring points. A target is connected to each target measuring point. The target is a corner reflector, a spherical reflector, a metal strip target, or a magnetically detachable target.

3. The method for monitoring static and dynamic parameters of a crane structure based on microwave sensing according to claim 1, characterized in that, In step 1.3, the field of view of the microwave rangefinder covers the entire range of all target measurement points reachable by the crane structure under all operating conditions.

4. The method for monitoring static and dynamic parameters of a crane structure based on microwave sensing according to claim 2, characterized in that, In step 2.1, under the background of strong metal reflection, multipath and clutter, the spatial resolution cell corresponding to the target or natural scattering point is locked by using prior distance window constraint, constant false alarm rate CFAR threshold detection and peak clustering algorithm. In step 2.2, Kalman filtering and / or α-β filtering are continuously performed on the spatial resolution cell where the target measurement point is located to achieve smooth updates and continuous locking of the target measurement point.

5. The method for monitoring static and dynamic parameters of a crane structure based on microwave sensing according to claim 1, characterized in that, In step 4.2, the deflection curve reconstruction employs at least one of spline fitting, polynomial fitting, shape function fitting that satisfies the structural boundary conditions, finite element prior shape function parameterization fitting, finite element digital twin online correction, and machine learning regression model to obtain the deflection curve distribution of the target measurement point.

6. The method for monitoring static and dynamic parameters of a crane structure based on microwave sensing according to claim 1, characterized in that, In step 5.2, the adaptive filter cutoff frequency is set in an engineering manner based on the structure's natural frequency range and the operation cycle to ensure that the quasi-static deflection component reflects the quasi-static deformation and the vibration deflection component reflects the vibration response. In step 5.2, the quasi-static / vibration separation algorithm employs at least one of bandpass filtering / moving average separation, wavelet decomposition, empirical mode decomposition (EMD), and variational mode decomposition (VMD) to obtain the quasi-static deflection component and the vibration deflection component.

7. The method for monitoring static and dynamic parameters of a crane structure based on microwave sensing according to claim 1, characterized in that, In step 5.2, based on the obtained quasi-static deflection components, the deterministic deflection components caused by the crane mechanism motion are also eliminated to obtain vibration residuals. The deterministic deflection components are obtained by constructing a deterministic model from at least one of the following operating conditions: amplitude angle, trolley position, lifting height, slewing angle, and running speed and its rate of change. The deterministic model is at least one of a regression model, a state-space model, or a parametric model based on mechanism kinematics / structural priors. The vibration residuals are the remaining components after deducting the quasi-static deflection components and the deterministic deflection components from the vertical deflection displacement time history. These vibration residuals are used as vibration deflection components for static / dynamic feature parameter extraction and dynamic amplification risk identification in step 4.

3.

8. The method for monitoring static and dynamic parameters of a crane structure based on microwave sensing according to claim 1, characterized in that, In step 6.1, duration criteria, confidence criteria, and multi-indicator consistency criteria are introduced to reduce the false alarm rate; In step 6.1, the crane's lifting capacity, lifting height, amplitude angle, trolley position, running speed, braking status and wind speed are accessed to establish a working condition label and differentiated threshold system. Different limits and criteria are used under different working condition labels to avoid false alarms caused by fixed thresholds. The online determination in step 6.1 also includes an anomaly identification algorithm, which triggers an alarm and outputs the anomaly type when the monitored value exceeds the limit or the anomaly continues to reach the preset condition; the anomaly type includes quasi-static deflection exceeding the limit, dynamic amplification anomaly, main frequency drift anomaly, and suspected trend stiffness degradation.

9. The method for monitoring static and dynamic parameters of a crane structure based on microwave sensing according to claim 1, characterized in that, In step 6.3, for any limit-type indicator The limit-type index Quasi-static deflection Vibration deflection peak-to-peak value or root mean square Dynamic amplification factor , main frequency frequency drift and trend volume Define the ratio of at least one of them. , Limit value indicator In working condition label The corresponding threshold; when And if the confidence criterion is met, output "hint"; when And if the duration criterion and consistency criterion are satisfied, output "warning"; when or However, if the duration reaches the second time threshold, an "alarm" will be output.

10. A monitoring system for static and dynamic parameters of a crane structure based on microwave sensing, characterized in that, include: A microwave rangefinder is used to emit electromagnetic waves and receive echo signals reflected from a target measuring point, so as to output at least one of the following information related to the target measuring point: distance information, velocity information, angle information, and echo intensity information. The installation support module is used to support and fix the microwave rangefinder, and to adjust the pitch and azimuth angles of the microwave rangefinder so that the microwave rangefinder beam covers the reachable space of the target measurement point under all working conditions. A target reflection enhancement module is installed at the target measurement point to enhance the signal-to-noise ratio and / or distinguishability of the echo signal, thereby improving the quality of the measurement point data. The working condition interface input is used to input at least one of the following working condition information: crane lifting capacity, lifting height, luffing angle, trolley position, running speed, braking status, and wind speed, and to generate working condition tags and start / stop / braking event segment information; The target retrieval and tracking module is used to acquire the echo signal and perform signal processing to achieve target detection and tracking of the target measurement point, and output a continuous measurement sequence of the target measurement point. The target retrieval and tracking module performs target detection based on at least one of prior distance window constraint, CFAR threshold detection, peak clustering, or peak retrieval. The target retrieval and tracking module uses Kalman filtering and / or α-β filtering to track and update the continuous measurement sequence. The prior distance window and / or angle window are dynamically updated according to the working condition information to adapt to the spatial position changes of the target measurement point caused by structural movement. The static parameter calculation module is used to calculate the camber, cantilever tilt, span, base distance and trolley gauge parameters of bridge and gantry cranes, as well as the verticality parameters of tower cranes, and the gauge, base distance, lifting range and amplitude of portal cranes, based on the line-of-sight distance of the spatial resolution unit where all target measurement points are located, and combined with the angle measurement information of the pitch angle and / or azimuth angle of the microwave rangefinder. The deflection calculation and reconstruction module is used to extract the relative displacement increment of the target measuring point in the line of sight direction based on the continuous measurement sequence, and to project the relative displacement increment into the vertical deflection displacement time history in the gravity direction by combining the pitch angle, azimuth angle and / or angle measurement information of the microwave rangefinder; and to reconstruct the deflection curve distribution of the crane structure based on the vertical displacement time history of multiple target measuring points. The quasi-static / vibration deflection separation and dynamic feature extraction module is used to obtain the quasi-static deflection component and static stiffness within segments of the same working condition label using a quasi-static / vibration separation algorithm, and to obtain the vibration deflection component by subtracting the quasi-static deflection component from the vertical deflection displacement time history; the quasi-static / vibration deflection separation and dynamic feature extraction module is also used to model and remove the deterministic deflection component caused by the motion of the crane mechanism to obtain the vibration residual, and to use the vibration residual as the vibration deflection component; the modeling of the deterministic deflection component is based on at least one of the working condition information of amplitude angle, trolley position, lifting height, slewing angle and running speed and its rate of change, and adopts at least one of regression model, state space model or parametric model based on mechanism kinematics / structural prior; and is used to calculate the peak-to-peak value, root mean square RMS, spectrum, frequency band energy and dynamic amplification factor (DAF) of the vibration deflection component; The threshold evaluation and early warning module is used to perform threshold evaluation based on at least one of the quasi-static deflection component, vibration deflection component, dynamic amplification factor DAF, main frequency drift and trend index, so as to output prompt, early warning or alarm information when deflection exceeds the limit, dynamic amplification is abnormal or trend stiffness degradation is suspected. The threshold assessment and early warning module selects differentiated thresholds and / or performs real-time threshold assessments based on the operating condition labels; It is also used for long-baseline trend modeling under the same working condition label to identify potential risks of loose connections, changes in support stiffness, or structural stiffness degradation.

Images