A steel bar truss floor support plate temporary support monitoring system based on acoustic vibration signals

Abstract

This invention discloses a monitoring system for temporary supports of steel truss floor slabs based on acoustic vibration signals, relating to the field of building engineering. It collects acoustic vibration response signals of the supports through standardized tapping and extracts multi-dimensional features. Innovatively, it combines real-time concrete pouring volume to calculate the dynamic load on the supports and utilizes a pre-trained machine learning model to generate dynamic diagnostic thresholds that change with the load, achieving adaptive adjustment of early warning standards. This accurately matches the actual stress risk of the supports, significantly improving the accuracy and foresight of hazard identification. It can promptly capture the progressive deterioration process from slight loosening to severe instability and triggers graded early warning and closed-loop response processes through a visual terminal. This deeply integrates advanced sensing technology with on-site safety management processes, effectively preventing safety accidents that may be caused by support failure and reducing emergency shutdowns through guided pre-maintenance, forming a new data-driven, traceable, and preventative safety management model.

Description

Technical Field

[0001] This invention relates to the field of building engineering, specifically to a monitoring system for temporary supports of steel truss floor slabs based on acoustic and vibration signals. Background Technology

[0002] In modern construction, especially for high-rise or large-span buildings, steel structures are widely used due to their high strength, lightweight, and excellent seismic performance. Floor decking is a precast component used to form building floors, combining the advantages of steel trusses and concrete. The stress process of floor decking is divided into the construction stage and the service stage. The construction stage is from the start of floor decking assembly until the concrete strength reaches the design requirements. During this process, the floor decking's self-weight, the weight of the wet concrete, and the construction live load are all borne by the steel truss, which is typically made of welded steel bars. Once the concrete strength reaches the design requirements, the floor decking enters the service stage. In this stage, the upper and lower chords of the steel truss work together with the concrete, and its stress performance and load-bearing capacity are the same as ordinary cast-in-place concrete floor slabs.

[0003] Steel truss floor decking, by processing steel bars into steel trusses, has high rigidity and can withstand its own weight and construction loads within a certain span. However, when the span is large, large deflection inevitably occurs at the mid-span, which can exceed the allowable value specified in the code (usually 1 / 180 or 20mm of the span). This not only affects the flatness of the bottom of the slab after pouring but also affects the structural safety due to plastic deformation. To ensure the overall strength, rigidity, and stability of the floor decking, temporary supports need to be installed under the floor decking after installation and before concrete pouring. The temporary supports and the floor decking work together as a whole to bear the self-weight of the component and construction loads. Therefore, the reliability between the floor decking and the temporary supports is the core of safety control during the construction phase.

[0004] Currently, in existing construction practices, we have found that the stability monitoring of these critical temporary supports mainly relies on traditional manual methods. Inspectors assess stability by visually observing the verticality of the supports or by feeling their movement. These methods are highly dependent on personal experience, subjective, difficult to quantify, and prone to missing checks in noisy and busy pouring sites. More importantly, existing technology lacks dynamic consideration of the actual load the supports bear. As concrete is continuously poured, the strength required for temporary supports gradually increases. This could lead to situations where the temporary supports are strong enough initially before concrete pouring, but become dangerous later during the continuous pouring of concrete.

[0005] Therefore, there is an urgent need for an intelligent monitoring technology that can provide real-time quantitative diagnosis and dynamically adjust early warning thresholds based on load, so as to upgrade safety management from passive inspection to proactive prevention. Summary of the Invention

[0006] The purpose of this invention is to provide a monitoring system for temporary supports of steel truss floor slabs based on acoustic vibration signals, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: 1. A monitoring system for temporary supports of steel truss floor slabs based on acoustic vibration signals, comprising a data acquisition module, an edge computing analysis module, and a visual monitoring and early warning terminal, characterized in that:

[0008] The data acquisition module is used to acquire the transient response signal of the temporary support under standardized impact excitation, and the transient response signal includes vibration signal and sound signal;

[0009] The edge computing analysis module is used to receive the transient response signal, extract feature vectors representing the contact state between the temporary support and the floor slab from the vibration signal and sound signal, calculate the dynamic construction load of the temporary support based on the real-time concrete pouring volume, generate dynamic signal feature thresholds corresponding to the current load level based on a pre-trained benchmark model of the correlation between construction load and contact state, compare and analyze the extracted feature vectors with the corresponding dynamic signal feature thresholds, and output the graded diagnosis results of the contact state.

[0010] The visualization monitoring and early warning terminal is used to display the support status, receive the diagnostic results output by the edge computing analysis module, and trigger corresponding early warning information and closed-loop response processes according to the risk level.

[0011] The system establishes reference signal characteristics for each support through initial calibration, and dynamically adjusts the diagnostic threshold according to load changes during concrete pouring, thereby achieving periodic and progressive proactive safety monitoring and early warning.

[0012] Preferably, the data acquisition module includes a striking unit and a signal acquisition sensor unit. The striking unit includes a built-in force sensor and a controller for applying standardized transient excitation to the temporary support at a preset position. The controller is configured to record and verify the force-time history of each strike to ensure that the excitation impulse is within a preset range and to issue a retest prompt for unqualified strikes.

[0013] Preferably, extracting feature vectors representing the contact state between the temporary support and the floor deck from the vibration and sound signals includes performing time-frequency domain analysis. The time-frequency domain analysis includes wavelet decomposition of the signals, calculating the energy ratio at different scales, and using the energy ratio at each scale as part of the feature vector, thereby fusing multi-dimensional information from the vibration and sound signals to generate feature vectors for condition diagnosis.

[0014] Preferably, the method for calculating the dynamic construction load of temporary supports based on the real-time obtained concrete pouring volume includes the following steps:

[0015] The location and number of each support point are determined according to the support layout diagram, and the influence area of ​​each support point is calculated according to the regional influence distribution method.

[0016] Receive real-time pouring data and calculate the weight of concrete acting on the area of ​​influence of the support point at the current moment based on the area of ​​influence.

[0017] By combining the self-weight of the floor deck with the construction live load, the real-time dynamic construction load value of the support point is generated.

[0018] Preferably, the establishment of the pre-trained benchmark model of associated construction load and contact state includes the following steps:

[0019] For supports and floor decks of the same type as those on site, a series of loads ranging from zero to above the design value were applied, and multiple contact states were simulated under each load.

[0020] Standardized tapping was performed under each combination of load and contact conditions. Vibration and sound signals were collected and feature vectors were extracted to form a training dataset.

[0021] A supervised learning algorithm is used to train a model, with load values ​​and signal feature vectors as inputs and the corresponding contact safety state as the output.

[0022] Preferably, the generation of the dynamic signal feature threshold includes: inputting the currently calculated dynamic construction load value into the benchmark model, and the benchmark model calling the feature space classification boundary corresponding to the load level according to the load value, and the classification boundary constitutes the dynamic signal feature threshold for on-site diagnosis.

[0023] Preferably, the contact state grading diagnosis and determination logic includes at least one of the following:

[0024] When all key features are within the safety threshold and the overall status score meets the preset conditions, it is judged to be safe.

[0025] When some key features reach the warning threshold, or when there is uncertainty in the overall status score, a warning is issued.

[0026] When the key characteristics significantly exceed the alarm threshold, or when the comprehensive status score clearly points to high risk, it is judged as dangerous;

[0027] The determination of the level is based not only on whether the change of the feature vector relative to its baseline value exceeds the threshold, but also on the dynamic adjustment in combination with the current dynamic construction load, so that the same change of feature can trigger a higher risk level under high load conditions.

[0028] Preferably, the visualization monitoring and early warning terminal also includes an initial calibration process, which includes: performing an initial tapping test on the temporary support, recording and storing the initial response signal of each support as a reference signal;

[0029] The support prompts for signal feature anomalies initially screened by the system are manually checked and verified. Based on the verification results, the baseline feature vectors of each support are determined or updated, and an initial calibration report is generated.

[0030] Preferably, the visualization monitoring and early warning terminal also includes dynamic construction loads of each support point calculated in real time based on the edge computing analysis module. During the concrete pouring process, when the load reaches the predetermined critical load stage threshold, a periodic monitoring task for the corresponding support is automatically generated and pushed to the detection terminal. The critical load stage threshold is set according to a preset proportion of the total design load.

[0031] Preferably, the closed-loop response process includes logging all early warning and alarm events throughout the entire process and automatically generating event reports. The event reports contain all data, timelines, and on-site records from anomaly discovery, early warning issuance, on-site handling to final verification and confirmation, which are used for post-event analysis and optimization of diagnostic models and construction processes.

[0032] In summary, the beneficial effects of this invention are:

[0033] This invention achieves a fundamental shift in the assessment of temporary support safety status from passive, discrete, and experience-based judgment to proactive, continuous, and quantitative early warning by integrating active excitation, multimodal signal analysis, and load-based dynamic intelligent diagnosis. It dynamically adjusts the judgment threshold using load values ​​calculated in real-time with the concrete pouring volume, ensuring a precise match between the early warning standards and the actual stress risk of the support. This significantly improves the accuracy and foresight of hazard identification, enabling timely capture of the progressive deterioration process from slight loosening to severe instability. Simultaneously, its "health fingerprint" calibration, periodic monitoring, intelligent graded early warning, and closed-loop response process deeply integrate advanced sensing technology with on-site safety management processes. This not only effectively prevents safety accidents that may be caused by support failure but also reduces emergency downtime through guided pre-maintenance, forming a data-driven, traceable, and preventative safety management model. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of the invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 This is a schematic diagram of the overall process framework of a monitoring system for temporary support of steel truss floor slabs based on acoustic vibration signals according to the present invention.

[0036] Figure 2 This is a schematic diagram of the flow frame structure of the edge computing module in a monitoring system for temporary support of steel truss floor slabs based on acoustic vibration signals according to the present invention.

[0037] Figure 3 This is a schematic diagram of the structure of temporary support and floor deck in a monitoring system for temporary support of steel truss floor deck based on acoustic vibration signals according to the present invention.

[0038] Figure 4 This is a schematic diagram of the floor deck structure in a monitoring system for temporary support of steel truss floor decks based on acoustic vibration signals, according to the present invention.

[0039] Figure 5 This is a schematic diagram of the finished floor deck structure in the monitoring system for temporary support of steel truss floor deck based on acoustic vibration signals according to the present invention. Detailed Implementation

[0040] The present invention will now be described in further detail with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention. These drawings are simplified schematic diagrams, which are only used to illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0041] To facilitate understanding of the present invention, a more complete description of the invention will be given below with reference to the accompanying drawings, which illustrate several embodiments of the invention. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that the disclosure of the invention will be more thorough and complete.

[0042] All features disclosed in this specification, or all steps in all disclosed methods or processes, may be combined in any way, except for mutually exclusive features and / or steps.

[0043] Any feature disclosed in this specification (including any appended claims, abstract, and drawings) may be replaced by other equivalent or similar features for a similar purpose, unless specifically stated otherwise. That is, unless specifically stated otherwise, each feature is merely one example of a series of equivalent or similar features.

[0044] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection, a direct connection, or an indirect connection through an intermediate medium; they can refer to the internal communication of at least two elements or the interaction relationship of at least two elements, unless otherwise explicitly limited. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0045] Please see Figure 1 - Figure 5 The present invention provides an embodiment of a monitoring system for temporary supports of steel truss floor slabs based on acoustic vibration signals. By actively striking the temporary support components, the system collects their vibration and acoustic response signals, analyzes the characteristics using signal processing and machine learning algorithms, and compares them with a dynamic threshold calculated based on the concrete pouring volume. This enables real-time diagnosis and graded early warning of the contact state between the temporary support and the floor slab. Specifically, the system includes a data acquisition module, an edge computing analysis module, and a visualization monitoring and early warning terminal.

[0046] The data acquisition module is used to acquire high-fidelity, repeatable transient response signals rich in support-contact state information, primarily using vibration signals with supplementary sound signals. Specifically, it includes:

[0047] The striking unit can be a standardized small hammer with a built-in force sensor. The force-time history of each strike can be recorded to ensure that the excitation energy is within a preset range, such as an impulse of 5-10 N·s. Marking points are set at the same height position of the temporary support for striking to avoid different modes being excited due to different striking points.

[0048] After each tap, the system automatically checks the peak value and duration of the excitation signal, and will prompt a retest for any taps that do not meet the requirements.

[0049] The sensor unit is used to collect vibration and sound signals. The vibration signal is installed in the middle of the temporary support, which is convenient to install and relatively less affected by operational interference, representing a good compromise between engineering practicality and signal sensitivity. The vibration wave, transmitted from the contact surface, already carries system path information.

[0050] Parameter selection:

[0051] Frequency response range ≥ 5kHz: The tapping pulse contains rich high-frequency components. High-frequency components are very sensitive to micro-slippage and local collisions (point contact) on the contact surface. Wide bandwidth ensures that key information is not lost. The vibration amplitude may be very small, especially in a tight state, requiring the sensor to be able to detect weak vibrations.

[0052] The sound signal, which collects the percussion sound and structural radiation sound propagating through the air, has a spectrum that is related to but not exactly the same as the vibration spectrum, and can provide additional, complementary features. It is non-contact, flexible in deployment, and can be used for rapid scanning or for detecting supports where it is difficult to install vibration sensors.

[0053] The construction site environment is noisy (pump trucks, vibrators). By using a directional microphone focused on the impact point and combined with synchronous triggering technology, data is collected only at the moment of impact, which can effectively suppress background noise.

[0054] Signal feature extraction is performed on the collected vibration and sound signals. From the original, high-dimensional waveform data, low-dimensional, robust feature vectors that characterize the physical state of the system (stiffness, damping, contact quality) are extracted for subsequent use, including:

[0055] 1. Time-domain characteristics: directly reflect energy and damping;

[0056] Peak amplitude: The acceleration value of the initial vibration peak. It directly reflects the local dynamic stiffness under impact. The looser the contact, the smaller the stiffness, and the peak amplitude may decrease (energy is absorbed by the loose contact surface) or a secondary impact peak may appear.

[0057] The decay time is a direct measure of the overall damping characteristics. When the contact is poor, energy dissipation is slow and the decay time becomes significantly longer.

[0058] Waveform factor & impulse factor: These are measures of the "sharpness" or "impact" of a waveform. They are highly sensitive to whether the signal contains impulsive components, such as intermittent impacts caused by looseness.

[0059] 2. Frequency Domain Characteristics (FFT): Reveals the inherent properties of the system.

[0060] Fundamental frequency (or dominant frequency): The first natural frequency of the system. This is one of the most crucial characteristics. System stiffness is directly proportional to the natural frequency; loose contact leads to a decrease in overall system stiffness, causing the fundamental frequency to drift significantly to lower frequencies.

[0061] Center of gravity frequency: the average position of the spectral energy distribution. As the high-frequency components decrease, representing the reduction of short-wavelength vibrations in rigid contact, the center of gravity frequency also shifts downward.

[0062] Frequency band energy distribution: The frequency spectrum is divided into multiple sub-bands, such as 0-500Hz, 500-1500Hz, and 1500-5kHz, and the energy proportion of each sub-band is calculated. Changes in contact state will alter the frequency distribution pattern of vibration energy.

[0063] Spectral entropy: measures the "disorder" or "complexity" of the spectrum. A clear, sharp resonant peak has a low spectral entropy value; when the contact is poor, the signal is cluttered, the bandwidth is broadened, and multiple clutter peaks may appear, leading to an increase in the spectral entropy value.

[0064] 3. Time-frequency domain features (wavelet transform): capturing non-stationary transient information

[0065] The signal is decomposed into N-level wavelet coefficients to obtain approximation coefficients and detail coefficients at different frequency scales;

[0066] Calculate the proportion of energy at each scale (frequency band) to the total energy. For example, a decrease in the energy proportion of scales representing high-frequency details may mean that the contact surface cannot effectively transmit high-frequency vibrations (i.e., poor contact).

[0067] These scale-energy ratios constitute a feature vector that is extremely sensitive to the contact state.

[0068] The edge computing analysis module includes building a dynamic threshold setting model based on concrete pouring volume. Its goal is to make the judgment standard for temporary supports no longer a fixed value, but an intelligent threshold that dynamically changes with the actual bearing risk of the support point, thereby achieving more accurate and reasonable early warnings. Specifically, it includes the following steps:

[0069] Step 1: Calculate the dynamic construction load at each support point.

[0070] Input data source:

[0071] Design data: Floor decking type and weight; steel truss weight.

[0072] Construction data:

[0073] Support layout diagram: a unique number for each support and its precise coordinates on the plane.

[0074] Concrete pouring plan: zoning, layering, pouring sequence and rate.

[0075] Real-time pouring volume: The actual volume of concrete poured to a certain support point area can be obtained by connecting with the concrete mixer truck or pump truck system, or by manual input by the construction worker.

[0076] Load distribution model:

[0077] The weight of the concrete poured on a floor slab is distributed to various support points. We use the "regional influence distribution method" to simplify the floor slab into a grid composed of support points, with each support bearing the load within its "influence area".

[0078] Determining the area of ​​influence: For regularly arranged supports, the "Vinnon polygon" principle is usually adopted. Simply put, the area of ​​influence of each support is the area enclosed by the perpendicular bisectors of the lines connecting it to adjacent supports. In a regular rectangular arrangement, the area of ​​influence of the middle support is the rectangular area formed by the support spacing.

[0079] Dynamic load calculation:

[0080] P i (t)=C*V i (t)+A i Q d +A i Q L

[0081] P i (t): The load borne by support point i at time t.

[0082] C: Unit weight of wet concrete (take 25-26 kN / m³).

[0083] V i (t): At time t, the area A affected by the support point i. i The volume of poured concrete within. This is a dynamic variable that changes over time.

[0084] Q d： Self-weight per unit area of ​​floor decking (including steel truss).

[0085] Q L： Construction live load (take the standard value, such as 1.5-2.0 kPa).

[0086] Result: The system can generate a load-time curve for each support point, reflecting the stress changes during the pouring process in real time.

[0087] Step 2: Establish a baseline database of "load-contact state-signal characteristics" (model training)

[0088] Full-condition test design:

[0089] Select the same type of support and floor decking as the site to build the test bench.

[0090] Simulate different loads: Apply a series of loads from 0 to the design limit load (e.g., 150%) to the support using jacks or counterweights.

[0091] Simulate different contact states:

[0092] Tight: The top support is completely tight with no gaps.

[0093] Slight loosening: Pre-set small gaps of 0.5mm, 1mm, etc.

[0094] Severe loosening / eccentricity: Simulates working conditions such as the top support not being leveled and there being debris at the bottom.

[0095] Data acquisition: Under each combination of "(load, contact state)", multiple standardized taps are performed to collect complete vibration and sound signals.

[0096] Feature library construction and model training:

[0097] Signal features are extracted from all collected vibration and sound signals to generate a large feature dataset.

[0098] Training is performed using supervised learning algorithms such as random forests, gradient boosting machines, or support vector machines.

[0099] Input: Signal feature vector (such as dominant frequency, decay time, spectral entropy, etc.) + the currently applied load value P.

[0100] Output / Label: Classification of contact status, such as: 0 - Safe, 1 - Warning, 2 - Danger.

[0101] After training, the algorithm has essentially learned the multidimensional feature space boundary that distinguishes different safety levels under any given load P. This boundary is the dynamic threshold.

[0102] Step 3: On-site dynamic threshold application and adaptive analysis

[0103] Threshold call:

[0104] When the on-site inspector taps the support numbered i, the testing terminal or cloud system will:

[0105] Based on the support's number, query its current real-time load P. i (t).

[0106] Load value P i (t) is input into the pre-trained diagnostic model.

[0107] Based on this load, the model automatically retrieves the characteristic threshold criteria applicable to this load level. For example, the allowable range of frequency drop under low loads will be more lenient than under high loads.

[0108] Intelligent diagnosis and early warning:

[0109] The system compares the feature vector extracted from the actual impact test with the threshold boundary under the corresponding load calculated by the model.

[0110] It not only provides a binary judgment of "loose / tight," but also performs risk classification:

[0111] Low load and slightly abnormal characteristics: may be marked as "Observation" to draw attention.

[0112] High load with slightly abnormal characteristics: may be directly upgraded to "warning" because the risk factor is amplified with the load.

[0113] Regardless of the load size, any severely abnormal characteristics will trigger an alarm.

[0114] Traditional methods use a fixed standard to inspect all supports, which may be too lenient for large supports that are already under load, and too strict for small supports that are not under load. This model realizes that the "load-dependent inspection" system can identify supports that "are not exceeding the standard under the current load, but will inevitably exceed the standard as the subsequent pouring load increases." It can issue an early warning before the concrete is poured to the next stage, requiring early tightening, thus achieving true preventive safety intervention.

[0115] It should also be noted that the visualization monitoring and early warning terminal is a periodic, progressive proactive safety management cycle, which can be divided into four main stages: initial calibration, real-time monitoring, intelligent diagnosis and graded early warning and response, as detailed below:

[0116] Phase 1: Initial calibration, establishing a "health fingerprint";

[0117] Objective: To establish a unique, initial health baseline for each support before load is applied, eliminating the influence of installation process and individual micro-variations.

[0118] Timing: After all temporary supports are installed according to design requirements and the top support is initially tightened, before concrete pouring begins.

[0119] Operating procedures:

[0120] System initialization: Load the electronic diagram of the support layout for this area into the handheld terminal or system.

[0121] Initial full inspection: Inspectors use standardized tools to perform a tapping test on each temporary support in the area. The system automatically records the response signal of each support at this time, which is the "reference signal".

[0122] Feature extraction and storage: The system extracts the feature values ​​of each supporting reference signal, binds them with the support ID and location information, and stores them as a "reference feature vector".

[0123] Manual sampling and verification:

[0124] Based on the baseline signal, the system initially screens out a few support signals with "abnormal" or "critical" characteristics.

[0125] On-site engineers or foremen conduct manual physical inspections of these marked supports, such as checking verticality, tapping to listen to the sound, and checking whether the top support is tightly attached.

[0126] Calibration decision: If the installation is confirmed to be secure by manual inspection, this signal characteristic is received as the "personalized" health benchmark for this support in the project, and the database is updated; if the installation is confirmed to have problems by manual inspection, adjustments are made, and the benchmark is updated after the adjustment is performed again.

[0127] Output: A validated Initial Calibration Report signifies that all support is in place and the system has a reliable starting point for subsequent change detection.

[0128] Phase Two: Real-time Monitoring

[0129] Objective: To periodically capture changes in the state of the support system during dynamic load increases and to promptly identify potential hazards.

[0130] Monitoring is not continuous, but rather conducted at points where the load experiences significant jumps, to ensure efficient and targeted monitoring. Typically, full-area or focused-area monitoring is performed at the following three key points:

[0131] When the concrete is poured to 1 / 3 of the design load: At this point, the support has already borne part of the load, and initial installation defects (such as loose connections) may begin to appear.

[0132] When the concrete is poured to 2 / 3 of the design load: This is one of the stages when the support is under the greatest stress and is a critical period to check whether its working condition is normal.

[0133] When the pouring is completed (full load): confirm whether all supports are still in a safe state under the final design load.

[0134] Monitoring Implementation: The cloud platform automatically pushes a list of monitoring tasks to the handheld terminals of the inspectors based on the pouring progress.

[0135] Following the task instructions, the inspector sequentially tapped the target supports to perform tests. The system automatically compared the new signals with the initial reference signals for that support.

[0136] After data collection, the system immediately performs a self-test, such as checking whether the signal-to-noise ratio and impact force meet the standards. If they do not meet the standards, the system prompts for a retest.

[0137] Phase 3: Intelligent Diagnosis

[0138] Feature calculation and change calculation: Calculate the rate of change or absolute change of each feature value of the current signal relative to its own reference value;

[0139] Dynamic threshold loading: The system loads the corresponding diagnostic threshold matrix based on the real-time calculated load currently borne by the support.

[0140] Multi-feature fusion diagnosis:

[0141] The system does not make judgments on each feature individually. Instead, it uses a pre-trained classification model (such as random forest) to input the current feature vector and the load value together. The model outputs a comprehensive state score or probability distribution, for example: [Safe: 85%, Warning: 12%, Danger: 3%].

[0142] Status determination:

[0143] Safety (Green): High overall score, and all key features are within the safety threshold.

[0144] Warning (yellow): If the overall score shows uncertainty, or 1-2 key features reach the warning threshold, the system may prompt "The main frequency has dropped by about 8%, and a re-examination is recommended".

[0145] Alarm (Red): The overall score indicates danger, or the key characteristics are seriously exceeded, and the system judges it as high risk.

[0146] Phase Four: Tiered Early Warning and Closed-Loop Response (From "Insight" to "Action")

[0147] Objective: To ensure that diagnostic results trigger clear, effective, and traceable on-site actions.

[0148] Tiered response mechanism:

[0149] Level 1: Warning (Yellow)

[0150] System Actions: Mark the support in yellow on the BIM / plan drawing; send push notifications to the mobile terminals of the area foreman and inspectors, including the support ID, location, and anomaly characteristics, such as "Support 3-B12, frequency dropped by 8%".

[0151] On-site response requirements: The foreman must arrange for personnel to manually inspect, fine-tighten, and reinforce the support within the specified time, and fill in the processing results in the system, and the system status will be changed to "pending review".

[0152] Level 2: Alarm (Red)

[0153] System actions:

[0154] Interface enhancement: On the visualization platform, this support flashes red and pops up a window.

[0155] On-site alarm: Automatically triggers nearby audible and visual alarms, emitting a continuous alarm.

[0156] Information push: Immediately send the highest priority alarm information to the project manager, safety director, and regional foreman, and automatically start voice calls or walkie-talkie broadcasts.

[0157] Process intervention: Optional - The system sends a pause signal to the concrete pumping control system to stop the continued pouring of the area.

[0158] On-site response requirements:

[0159] Work must be stopped immediately: On-site workers must suspend operations in this area and adjacent areas.

[0160] Emergency response: Technicians and safety officers should immediately rush to the scene and, based on the possible causes indicated by the system (such as "severe eccentricity"), carry out reinforcement, jacking, or replacement of supports.

[0161] Closed-loop acceptance: After the handling is completed, the support must be re-inspected until the system test results return to a "safe" state. Only after confirmation by the safety supervisor can the alarm be deactivated and construction resume.

[0162] Data backtracking and optimization:

[0163] All alarms, warnings, and their handling processes are automatically logged.

[0164] Regularly analyze early warning / alarm events to optimize threshold models, improve support installation processes, and form a closed loop for continuous improvement of safety management.

[0165] Here is a specific operational example, which is a commercial complex project. The standard floor uses a steel truss floor slab with a span of 9 meters. In order to control the deflection during the construction phase, a row of temporary independent steel supports with a spacing of 2 meters is set in the middle of the floor slab. Taking the temporary independent steel support numbered B-12 as an example;

[0166] 1. Initial calibration phase: Establishing a "health fingerprint"

[0167] Time: The afternoon before the concrete pouring.

[0168] The inspector conducted the first calibration test on the support numbered B-12 by performing standardized tapping at the designated location.

[0169] Data and Results:

[0170] The system records the reference signal of this tap.

[0171] Key features extracted: main frequency F_baseline = 185 Hz, decay time Td_baseline = 0.25 seconds.

[0172] The system initially determined that the signal was "clear". At the same time, based on the support layout diagram, the system automatically calculated the affected area of ​​support B-12 as 2m x 2m = 4㎡ and stored it.

[0173] The system randomly selects B-12 for review. On-site inspection confirms that the top support is tightly fitted to the floor slab and has good verticality. The system then clicks "Confirm Pass" in the app.

[0174] Output: B-12's "health fingerprint" -- {Support ID: B-12, Base frequency: 185Hz, Base decay: 0.25s, Status: Healthy} is stored in the database.

[0175] 2. Real-time monitoring and dynamic threshold setting stage: The threshold is adjusted accordingly as the load changes.

[0176] Time: Concrete pouring is in progress, and has reached approximately 1 / 3 of the thickness of this area (equivalent to a load of approximately 4 kPa).

[0177] System background calculation:

[0178] Real-time load: Current load P of B-12 support = (4 kPa concrete + 0.3 kPa floor slab + 1.5 kPa construction live load) * 4㎡ = 23.2 kN.

[0179] Model input: The system inputs a load of 23.2 kN into the dynamic threshold model.

[0180] Generate dynamic threshold: The model outputs the decision threshold for B-12 under the current load.

[0181] Warning threshold: Frequency drop >10%, or decay time increase >40%.

[0182] Alarm threshold: frequency drop >20%, or decay time increase >80%.

[0183] 3. First periodic monitoring: detecting any abnormal signs.

[0184] Time: Approximately 30 minutes after pouring to 1 / 3 of the load.

[0185] Operation: The employee receives a system task to retest B-12.

[0186] Data and Diagnostics:

[0187] The detector acquires a new signal, and the system automatically calculates the characteristics: current main frequency F_current = 168 Hz, current attenuation time Td_current = 0.31 seconds.

[0188] Calculate the change:

[0189] The rate of change of the dominant frequency ΔF = (168-185) / 185 = -9.2%

[0190] The decay time change ΔTd = 0.31 - 0.25 = 0.06 seconds (an increase of 24%).

[0191] Intelligent Judgment: The system inputs {ΔF=-9.2%, ΔTd=+24%, load=23.2kN} into the classifier. Judgment Result: Warning (yellow) is issued because ΔTd (24%) has exceeded half of the warning threshold (40%), and the main frequency is dropping close to the warning line, indicating an increased overall risk probability.

[0192] Early warning and response:

[0193] System: On the project's monitoring screen, the B-12 icon flashes yellow. Employees receive a notification: "Warning: Support B-12 has an abnormally increased attenuation time of 24%, which may indicate a slight loosening. Please recheck."

[0194] On-site: The employee paused his work and went to location B-12. Inspection revealed that due to vibrations from the concrete pouring, the locking handle of the support had slightly slipped, causing a slight loss of preload on the top support. He immediately used a wrench to tighten it again.

[0195] Closed loop: The employee uploaded the processed photo with the note "Re-tightened". The system status changed to "Processed, pending review".

[0196] 4. Secondary monitoring and crisis early warning: Escalation of hidden dangers

[0197] Time: When the concrete is poured to 2 / 3 of its thickness (equivalent to a load of approximately 8 kPa).

[0198] System backend: The load of B-12 is updated to P = 38.4 kN, and the alarm threshold called by the system is tightened accordingly (for example, an alarm is triggered if the main frequency drops by more than 15%).

[0199] Operation: Employees conduct critical node retesting on B-12.

[0200] Data and Diagnostics:

[0201] Characteristic values: F_current = 152 Hz, Td_current = 0.52 seconds.

[0202] Changes: ΔF = -17.8%, ΔTd = +108%.

[0203] Judgment: Alarm (Red)! Both the main frequency and decay time far exceed the alarm threshold under the current load. The classifier outputs a "dangerous" probability exceeding 95%.

[0204] Early warning and emergency response:

[0205] system:

[0206] On the monitoring screen, B-12 turned red and flashed violently, and a huge alarm box popped up.

[0207] The on-site audible and visual alarm sounded, displaying the message: "Emergency! Support B-12 is at risk of instability! Possible causes: severe loosening or eccentricity. It is recommended to immediately stop pouring and take appropriate action!"

[0208] On-site (Emergency Response):

[0209] Upon hearing the alarm and seeing the hand signals, the concrete pump truck operator immediately stopped delivering concrete to the area.

[0210] Upon inspection, it was found that because the previous thread slippage problem had not been completely resolved, the top support tilted significantly under the continuously increasing load, and the support was under eccentric compression.

[0211] The emergency response team immediately took action: an emergency backup support was installed next to B-12, and the damaged B-12 support was unloaded and replaced.

[0212] Closed loop and review:

[0213] After replacement, the staff tested the new support, and the signal characteristics returned to normal, with the system displaying "safe".

[0214] Once the alarm was lifted and the project manager and safety director jointly confirmed the resumption of pouring, the process was resumed.

[0215] Afterwards, the system automatically generated the "B-12 Support Alarm Event Report", which included all data, timelines and photos from the first warning to the final handling. The project meeting reviewed this and decided to conduct a general inspection and replacement of the locking handles of all supports in the same batch.

[0216] The above description is merely a specific embodiment of the invention, but the scope of protection of the invention is not limited thereto. Any variations or substitutions conceived without inventive effort should be included within the scope of protection of the invention. Therefore, the scope of protection of the invention should be determined by the scope defined in the claims.

Claims

1. A monitoring system for temporary supports of steel truss floor slabs based on acoustic and vibration signals, comprising a data acquisition module, an edge computing analysis module, and a visual monitoring and early warning terminal, characterized in that: The data acquisition module is used to acquire the transient response signal of the temporary support under standardized impact excitation, and the transient response signal includes vibration signal and sound signal; The edge computing analysis module is used to receive the transient response signal, extract feature vectors representing the contact state between the temporary support and the floor slab from the vibration and sound signals, and calculate the dynamic construction load of the temporary support based on the real-time concrete pouring volume. This includes determining the position and number of each support point according to the support layout diagram and calculating the influence area of ​​each support point according to the regional influence distribution method. It also receives input real-time pouring data, calculates the weight of concrete acting on the influence area of ​​the support point at the current moment based on the influence area, and generates a real-time dynamic construction load value for the support point by combining the self-weight of the floor slab and the construction live load. Based on a pre-trained benchmark model that correlates construction load and contact state, it generates a dynamic signal feature threshold corresponding to the current load level. The generation of the dynamic signal feature threshold includes: inputting the currently calculated dynamic construction load value into the benchmark model; the benchmark model calling the feature space classification boundary corresponding to the load level based on the load value; the classification boundary constituting the dynamic signal feature threshold for on-site diagnosis; comparing and analyzing the extracted feature vector with the corresponding dynamic signal feature threshold; and outputting the graded diagnosis result of the contact state. The visualization monitoring and early warning terminal is used to display the support status, receive the diagnostic results output by the edge computing analysis module, and trigger corresponding early warning information and closed-loop response processes according to the risk level. 2.The steel bar truss floor support monitoring system based on acoustic vibration signal according to claim 1, characterized in that: The data acquisition module includes a striking unit and a signal acquisition sensor unit. The striking unit includes a built-in force sensor and a controller for applying standardized transient excitation to the temporary support at a preset position. The controller is configured to record and verify the force-time history of each strike to ensure that the excitation impulse is within a preset range and to issue a retest prompt for unqualified strikes.

3. The system according to claim 2, wherein the system is characterized by: Extracting feature vectors representing the contact state between temporary supports and floor decking from the vibration and sound signals includes performing time-frequency domain analysis. The time-frequency domain analysis includes wavelet decomposition of the signals, calculating the energy ratio at different scales, and using the energy ratio at each scale as part of the feature vector, thereby fusing multi-dimensional information from vibration and sound signals to generate feature vectors for condition diagnosis.

4. The monitoring system for temporary supports of steel truss floor slabs based on acoustic vibration signals according to claim 3, characterized in that: The establishment of the pre-trained benchmark model of associated construction loads and contact states includes the following steps: For supports and floor decks of the same type as those on site, a series of loads ranging from zero to above the design value were applied, and multiple contact states were simulated under each load. Standardized tapping was performed under each combination of load and contact conditions. Vibration and sound signals were collected and feature vectors were extracted to form a training dataset. Supervised learning algorithms are used to train a model, with load values ​​and signal feature vectors as inputs and the corresponding contact safety state as the output.

5. The monitoring system for temporary supports of steel truss floor slabs based on acoustic vibration signals according to claim 4, characterized in that: The grading diagnosis and determination logic for the contact state includes at least one of the following: When all key features are within the safety threshold and the overall status score meets the preset conditions, it is judged to be safe. When some key features reach the warning threshold, or when there is uncertainty in the overall status score, a warning is issued. When the key characteristics significantly exceed the alarm threshold, or when the comprehensive status score clearly points to high risk, it is judged as dangerous; The determination of the risk level is based not only on whether the change of the feature vector relative to its baseline value exceeds the threshold, but also on the dynamic adjustment of the current dynamic construction load, so that the same change of feature can trigger a higher risk level under high load conditions.

6. The monitoring system for temporary supports of steel truss floor slabs based on acoustic vibration signals according to claim 1, characterized in that: The visualization monitoring and early warning terminal also includes an initial calibration process, which includes: performing an initial tapping test on the temporary support, recording and storing the initial response signal of each support as a reference signal; The support prompts for signal feature anomalies initially screened by the system are manually checked and verified. Based on the verification results, the baseline feature vectors of each support are determined or updated, and an initial calibration report is generated.

7. The monitoring system for temporary supports of steel truss floor slabs based on acoustic vibration signals according to claim 6, characterized in that: The visualization monitoring and early warning terminal also includes dynamic construction loads of each support point calculated in real time based on the edge computing analysis module. During the concrete pouring process, when the load reaches the predetermined critical load stage threshold, a periodic monitoring task for the corresponding support is automatically generated and pushed to the detection terminal. The critical load stage threshold is set according to a preset proportion of the total design load.

8. The monitoring system for temporary supports of steel truss floor slabs based on acoustic vibration signals according to claim 7, characterized in that: The closed-loop response process includes logging all early warning and alarm events throughout the entire process and automatically generating event reports. The event reports contain all data, timelines, and on-site records from anomaly discovery, early warning issuance, on-site handling to final verification and confirmation, which are used for post-event analysis and optimization of diagnostic models and construction processes.

Images