Methods and systems for assessing compliance of safety behaviors at highway construction sites

Abstract

This invention relates to the field of construction safety monitoring technology, and provides a method and system for assessing the compliance of safety behaviors at highway construction sites. The method includes the following steps: real-time acquisition of ground coupled vibration data; removal of interference features from the ground coupled vibration data, extraction of structural vibration feature sequences and high-frequency micro-vibration feature sequences, and reconstruction of these sequences to obtain a set of mechanical operation behavior parameters and a set of gait behavior parameters; synchronous mapping of the mechanical operation behavior parameter set and the gait behavior parameter set to a unified three-dimensional geographic information spatiotemporal grid to generate a dynamic construction activity coupled field; and real-time determination of human-machine interaction distance and the risk of mechanical operation restricted area intrusion within the construction activity coupled field to obtain a safety behavior compliance assessment result. This invention can complete on-site safety behavior compliance determination without relying on optical line-of-sight, effectively overcoming the recognition deficiencies of traditional video surveillance under conditions such as dust, rain, fog, and nighttime.

Description

Technical Field

[0001] This invention belongs to the field of construction safety monitoring technology, and in particular relates to a method and system for assessing the compliance of safety behaviors at highway construction sites. Background Technology

[0002] Highway construction projects are characterized by long construction routes, dispersed work areas, multiple trades operating simultaneously, and dense operation of large machinery. Construction sites often involve complex interactions between various heavy machinery such as road rollers, pavers, material transport vehicles, and cranes, and a large number of construction workers. Frequent human-machine interaction, dynamically changing work areas, and numerous blind spots make highway construction sites high-risk areas for safety accidents. Among these accidents, the main ones are directly related to unsafe behaviors such as personnel entering restricted areas for machinery operation and insufficient safe distances between workers and machinery. Therefore, real-time compliance assessments of the safety behaviors of construction personnel and machinery are crucial for accident prevention.

[0003] Currently, traditional safety behavior assessments at highway construction sites mainly rely on manual inspections or fixed camera video analysis. However, existing technologies have the following shortcomings that urgently need improvement: First, video surveillance has limitations in line of sight and blind spots, and its recognition accuracy drops sharply in dusty, rainy, foggy, or low-light conditions at night; second, existing technologies are mostly based on single-point independent judgments, lacking a comprehensive assessment of the coupled dynamics between mechanical operation actions and personnel walking trajectories, making it difficult to detect deeper violations. Summary of the Invention

[0004] The purpose of this invention is to provide a method for assessing the compliance of safety behaviors at highway construction sites, thereby addressing the aforementioned technical problems.

[0005] This invention is implemented as follows: a method for assessing compliance with safety behaviors at highway construction sites, comprising the following steps:

[0006] Real-time acquisition of ground coupled vibration data; the ground coupled vibration data includes ground vibration signals collected in real time by multiple geophone nodes deployed at the construction site;

[0007] Interference features are removed from the ground coupled vibration data, and the structural vibration feature sequence induced by mechanical operation and the high-frequency micro-vibration feature sequence induced by human gait are extracted.

[0008] The structural vibration feature sequence and the high-frequency micro-vibration feature sequence are reconstructed to obtain a mechanical operation behavior parameter set and a gait behavior parameter set. The mechanical operation behavior parameter set includes the machine type, operation trajectory, and operation status. The gait behavior parameter set includes the gait event sequence, movement path, and gait rhythm features.

[0009] The mechanical operation behavior parameter set and the gait behavior parameter set are synchronously mapped to a unified three-dimensional geographic information spatiotemporal grid to generate a dynamic construction activity coupling field.

[0010] Based on preset safety behavior rules, the human-machine interaction distance and the risk of intrusion into the restricted area of ​​mechanical operation are determined in real time in the construction activity coupling field, and the safety behavior compliance assessment results are obtained.

[0011] Furthermore, the steps of removing interference features from the ground coupled vibration data and extracting the structural vibration feature sequence induced by mechanical operations and the high-frequency micro-vibration feature sequence induced by personnel gait specifically include:

[0012] Based on the beamforming algorithm, non-construction vehicle interference is removed from the ground coupled vibration data to obtain the first processed data;

[0013] Based on the energy distribution entropy method, environmental interference is identified, and adaptive parameter adjustments are made to the first processed data to obtain the second processed data.

[0014] Based on variational mode decomposition, structural vibration characteristic sequences and high-frequency micro-vibration characteristic sequences are separated from the second processed data.

[0015] Furthermore, the method for reconstructing the structural vibration characteristic sequence includes the following steps:

[0016] Extract the time-frequency energy peak envelope from the structural vibration feature sequence to identify the mechanical type;

[0017] The direction of arrival of the vibration characteristic sequence of the structure is estimated using an inversion algorithm, and the continuous position coordinates of the vibration source in the geographic coordinate system are calculated to form the operation trajectory.

[0018] Based on the repetition frequency and amplitude characteristics of the periodic impact pulses in the structural vibration characteristic sequence, the excitation state and cumulative number of compaction passes of the machinery are determined as the working state.

[0019] Furthermore, the method for reconstructing the high-frequency micro-vibration feature sequence includes the following steps:

[0020] Peak detection and pulse interval analysis are performed on the high-frequency micro-vibration feature sequence to extract the time series of gait impact events and construct a gait event sequence;

[0021] The arrival time difference of multiple points in the high-frequency micro-vibration feature sequence is calculated using the triangulation method, and the spatial coordinates corresponding to each gait impact event are solved to generate a movement path;

[0022] Based on the consistency of the interval duration between adjacent gait impact events in the gait event sequence, gait rhythm features are calculated to identify personnel identity and abnormal gait.

[0023] Furthermore, the step of synchronously mapping the set of mechanical operation behavior parameters and the set of gait behavior parameters onto a unified three-dimensional geographic information spatiotemporal grid to generate a dynamic construction activity coupling field specifically includes:

[0024] Construct a three-dimensional geographic information spatiotemporal grid covering the entire construction site;

[0025] The operation trajectory and operation status of the mechanical operation behavior parameter set are mapped to the three-dimensional geographic information spatiotemporal grid to generate a dynamically updated mechanical operation heat map.

[0026] The movement paths and gait event sequences in the gait behavior parameter set are mapped to the three-dimensional geographic information spatiotemporal grid to generate a dynamically updated personnel distribution density field.

[0027] By using timestamp alignment and spatial interpolation algorithms, the mechanical operation heatmap and the personnel distribution density field are fused to generate a dynamic construction activity coupling field.

[0028] Another objective of this invention is to provide an assessment system for compliance of safety behaviors at highway construction sites, used to implement the assessment method for compliance of safety behaviors at highway construction sites as described above, specifically including:

[0029] The vibration data acquisition module is used to acquire ground coupled vibration data in real time; the ground coupled vibration data includes ground vibration signals collected in real time by multiple geophone nodes deployed at the construction site.

[0030] The feature extraction module is used to remove interfering features from the ground coupled vibration data and extract the structural vibration feature sequence induced by mechanical operation and the high-frequency micro-vibration feature sequence induced by personnel gait.

[0031] The feature reconstruction module is used to reconstruct the structural vibration feature sequence and the high-frequency micro-vibration feature sequence respectively to obtain a mechanical operation behavior parameter set and a gait behavior parameter set; the mechanical operation behavior parameter set includes the machine type, operation trajectory and operation status; the gait behavior parameter set includes the gait event sequence, movement path and gait rhythm features;

[0032] The coupling field construction module is used to synchronously map the set of mechanical operation behavior parameters and the set of gait behavior parameters to a unified three-dimensional geographic information spatiotemporal grid to generate a dynamic construction activity coupling field.

[0033] The compliance assessment module is used to determine the human-machine interaction distance and the risk of intrusion into the restricted area of ​​mechanical operation in real time in the construction activity coupling field based on preset safety behavior rules, and obtain the safety behavior compliance assessment results.

[0034] Furthermore, the feature extraction module specifically includes:

[0035] The vehicle interference removal unit is used to remove non-construction vehicle interference from the ground coupled vibration data based on the beamforming algorithm to obtain the first processed data;

[0036] An environmental interference identification unit is used to identify environmental interference based on the energy distribution entropy method, and to adaptively adjust the parameters of the first processed data to obtain the second processed data.

[0037] The feature separation unit is used to separate the structural vibration feature sequence and the high-frequency micro-vibration feature sequence from the second processed data based on the variational mode decomposition method.

[0038] Furthermore, the feature reconstruction module specifically includes:

[0039] The mechanical type identification unit is used to extract the time-frequency energy peak envelope from the structural vibration feature sequence to identify the mechanical type.

[0040] The operation trajectory generation unit is used to estimate the direction of arrival of the vibration feature sequence of the structure using an inversion algorithm, calculate the continuous position coordinates of the vibration source in the geographic coordinate system, and form the operation trajectory.

[0041] The operation status determination unit is used to determine the excitation state and cumulative number of compaction passes of the machine based on the repetition frequency and amplitude characteristics of the periodic impact pulses in the structural vibration characteristic sequence, and to use this as the operation status.

[0042] Furthermore, the feature reconstruction module specifically includes:

[0043] The event sequence construction unit is used to perform peak detection and pulse interval analysis on the high-frequency micro-vibration feature sequence, extract the time series of gait impact events, and construct the gait event sequence.

[0044] The movement path generation unit is used to calculate the time difference of arrival of multiple points in the high-frequency micro-vibration feature sequence using the triangulation method, solve the spatial coordinates corresponding to each gait impact event, and generate a movement path.

[0045] The rhythm feature calculation unit is used to calculate gait rhythm features based on the consistency of the interval duration between adjacent gait impact events in the gait event sequence, and to identify personnel identity and abnormal gait.

[0046] Furthermore, the coupling field construction module specifically includes:

[0047] Spatiotemporal grid construction unit, used to construct a three-dimensional geographic information spatiotemporal grid covering the entire construction site;

[0048] The first mapping unit is used to map the operation trajectory and operation status of the mechanical operation behavior parameter set to the three-dimensional geographic information spatiotemporal grid to generate a dynamically updated mechanical operation heat map.

[0049] The second mapping unit is used to map the movement paths and gait event sequences in the gait behavior parameter set to the three-dimensional geographic information spatiotemporal grid, thereby generating a dynamically updated personnel distribution density field.

[0050] The human-machine fusion unit is used to fuse the mechanical operation heat map with the personnel distribution density field through timestamp alignment and spatial interpolation algorithms to generate a dynamic construction activity coupling field.

[0051] This invention provides a method for assessing the compliance of safety behaviors at highway construction sites. By extracting mechanical operation behavior parameter sets and gait behavior parameter sets from ground coupled vibration signals and constructing a dynamic construction activity coupling field in a three-dimensional geographic information spatiotemporal grid, it can complete the on-site safety behavior compliance judgment without relying on optical line of sight. This effectively overcomes the recognition defects of traditional video surveillance under conditions such as dust, rain, fog, and night, and significantly improves the robustness and intelligence level of safety behavior compliance assessment. Attached Figure Description

[0052] Figure 1 A flowchart illustrating the method for assessing the compliance of safety behaviors at highway construction sites, as provided in this embodiment of the invention.

[0053] Figure 2 This is a flowchart illustrating step S200 in the method for assessing the compliance of safety behaviors at highway construction sites provided in an embodiment of the present invention.

[0054] Figure 3 This is a flowchart illustrating the method for reconstructing structural vibration characteristic sequences provided in an embodiment of the present invention.

[0055] Figure 4 This is a flowchart illustrating the method for reconstructing high-frequency micro-vibration feature sequences provided in an embodiment of the present invention.

[0056] Figure 5 This is a flowchart illustrating step S400 in the method for assessing the compliance of safety behaviors at highway construction sites provided in an embodiment of the present invention.

[0057] Figure 6 A schematic diagram of the structure of the highway construction site safety behavior compliance assessment system provided in this embodiment of the invention.

[0058] Figure 7 This is a schematic diagram of the feature extraction module provided in an embodiment of the present invention.

[0059] Figure 8 This is a schematic diagram of the feature reconstruction module provided in an embodiment of the present invention.

[0060] Figure 9 This is a schematic diagram of the structure of the coupling field construction module provided in an embodiment of the present invention. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0062] like Figure 1 As shown, in one embodiment of the present invention, a method for assessing the compliance of safety behaviors at highway construction sites is provided, comprising the following steps:

[0063] S100. Real-time acquisition of ground coupled vibration data; the ground coupled vibration data includes ground vibration signals collected in real time by multiple geophone nodes deployed at the construction site;

[0064] S200. Remove interference features from the ground coupled vibration data and extract the structural vibration feature sequence induced by mechanical operation and the high-frequency micro-vibration feature sequence induced by personnel gait.

[0065] S300. The structural vibration feature sequence and the high-frequency micro-vibration feature sequence are reconstructed to obtain a mechanical operation behavior parameter set and a gait behavior parameter set; the mechanical operation behavior parameter set includes mechanical type, operation trajectory and operation status; the gait behavior parameter set includes gait event sequence, movement path and gait rhythm features;

[0066] S400. Synchronously map the set of mechanical operation behavior parameters and the set of gait behavior parameters to a unified three-dimensional geographic information spatiotemporal grid to generate a dynamic construction activity coupling field.

[0067] S500, based on preset safety behavior rules, determines the human-machine interaction distance and the risk of intrusion into the restricted area of ​​mechanical operation in real time in the construction activity coupling field, and obtains the safety behavior compliance assessment results.

[0068] Specifically, ground coupled vibration data can be acquired in real time through a distributed microelectromechanical system (MEMS) seismic detector array deployed at the highway construction site. In practical applications, multiple detector nodes are buried or deployed on the surface at a preset grid spacing (e.g., 10m x 10m) along the roadbed, pavement, and work area edges of the highway construction site. These detector nodes communicate wirelessly, forming a distributed sensing network covering the entire construction site. Each detector can sensitively acquire vibration acceleration signals from the ground medium, obtaining ground vibration signals with a sampling frequency of no less than 1000Hz to ensure complete capture of high-frequency micro-vibration signals from personnel gait. Each detector node synchronously acquires ground vibration signals during construction, forming ground coupled vibration data. This ground coupled vibration data specifically includes low-frequency, high-intensity structural vibrations generated by heavy machinery operations, high-frequency, low-intensity micro-vibrations generated by personnel walking, and environmental noise interference.

[0069] like Figure 2 As shown, in a preferred embodiment of the present invention, the steps of removing interference features from the ground coupled vibration data and extracting the structural vibration feature sequence induced by mechanical operations and the high-frequency micro-vibration feature sequence induced by personnel gait, i.e., step S200, specifically include:

[0070] S210. Based on the beamforming algorithm, non-construction vehicle interference is removed from the ground coupled vibration data to obtain the first processed data;

[0071] Specifically, the arrival time difference sequence of ground vibration signals from each geophone node is acquired, and the velocity vector of the vibration source corresponding to each geophone node within a continuous time window is calculated using a beamforming algorithm. When the magnitude of the velocity vector is continuously greater than a preset high-speed traffic threshold (e.g., 36 km / h), and the angle between the trajectory direction of the vibration source and the preset road axis direction is less than a preset angle (e.g., 15°), the ground vibration signal collected by the geophone node corresponding to that vibration source is marked as non-construction vehicle interference and removed from the original ground coupled vibration data to obtain the first processed data. This step effectively avoids interference from the passage of social vehicles or management vehicles on subsequent analysis.

[0072] S220. Based on the energy distribution entropy method, identify environmental interference, and adaptively adjust the parameters of the first processed data to obtain the second processed data.

[0073] In practical applications, environmental interference mainly refers to rainfall interference. Specifically, the short-time energy distribution entropy value is calculated for the first processed data. When the energy distribution entropy value of the signal is greater than the preset entropy threshold (indicating high signal randomness) within multiple consecutive time windows, and the signal frequency band is concentrated in the preset rainfall impact characteristic frequency band (such as 1-5kHz), the time period is marked as the rainfall interference period. During the rainfall interference period, the signal-to-noise ratio trigger threshold of the human gait characteristic frequency threshold is automatically increased, and the lower limit of the frequency band of the mechanical characteristic frequency threshold is appropriately reduced to suppress the interference of rainfall impact noise on the effective signal.

[0074] S230. Based on the variational mode decomposition method, the structural vibration characteristic sequence and the high-frequency micro-vibration characteristic sequence are separated from the second processed data.

[0075] Specifically, based on the differences in energy distribution and center frequency distribution characteristics of the ground vibration signals in the second processed data, the mixed ground vibration signals are adaptively decomposed into several intrinsic mode function components. Then, by setting mechanical characteristic frequency thresholds (e.g., below 50Hz) and human gait characteristic frequency thresholds (e.g., above 80Hz), components below the mechanical characteristic frequency threshold are assigned to the structural vibration characteristic sequence, while components above the human gait characteristic frequency threshold are assigned to the high-frequency micro-vibration characteristic sequence. It should be noted that wavelet packet transform can be performed on the separated structural vibration characteristic sequence and high-frequency micro-vibration characteristic sequence respectively to extract the energy features of each frequency band, forming feature vectors for subsequent mechanical behavior reconstruction and gait behavior reconstruction. Wavelet packet transform can provide fine resolution in both the time and frequency domains, effectively preserving the transient and periodic characteristics of the vibration signal.

[0076] like Figure 3 As shown, in a preferred embodiment of the present invention, the method for reconstructing the structural vibration characteristic sequence includes the following steps:

[0077] S311. Extract the time-frequency energy peak envelope from the structural vibration feature sequence to identify the mechanical type;

[0078] Specifically, firstly, a short-time Fourier transform is performed on the structural vibration characteristic sequence to convert the time-domain vibration signal into a time-frequency domain representation. The window length is set to 2048 sampling points, the window overlap rate to 50%, and the frequency resolution to 0.5Hz. The transformed signal generates a two-dimensional time-frequency spectrum, with time on the horizontal axis and frequency on the vertical axis. The grayscale value of each pixel represents the energy intensity at that time-frequency point. Next, peak detection is performed on the frequency domain energy of each time window along the time axis on the time-frequency spectrum. For each time window, several peak points with the highest energy on the frequency axis are extracted, and their frequency and energy amplitude are recorded. The peak points of consecutive time windows are connected in chronological order to form a time-frequency energy peak envelope curve. This envelope curve reflects the dynamic change of the main energy of mechanical vibration over time. In practical applications, vibration signals of various typical machines (such as road rollers, pavers, and material transport vehicles) under different operating conditions are collected in advance at highway construction sites. The time-frequency energy peak envelopes of each machine are extracted using the above method to establish a mechanical vibration feature library. The mechanical vibration feature library stores the following characteristic parameters of various machines, for example:

[0079] Roller characteristics: low-frequency periodic impact pulses, impact frequency is related to the speed of the roller (usually 8-15Hz), the time-frequency energy peak envelope shows a regular pulse peak group, the pulse interval is stable, and the spectral energy is mainly concentrated on the fundamental frequency and its harmonic components;

[0080] Paver characteristics: Continuous and stable broadband vibration signal, the time-frequency energy peak envelope shows a relatively smooth band-shaped distribution, the energy is mainly concentrated in the mid-low frequency range (20-80Hz), and there are no obvious pulse characteristics;

[0081] Material transport vehicle characteristics: intermittent low-frequency vibration, time-frequency energy peak envelope presents sparse isolated pulse peaks, pulse intervals are uneven, energy is concentrated in the low frequency range (10-40Hz), and the amplitude difference between no-load and full-load states is significant.

[0082] Then, the time-frequency energy peak envelope extracted in real time is matched with various mechanical features in the mechanical vibration feature library; the dynamic time warping algorithm is used to calculate the similarity between the envelope curve to be identified and each template curve in the feature library. Dynamic time warping can effectively handle the expansion and contraction deformation of the envelope curve on the time axis under different speed conditions; the mechanical type with the highest similarity is selected as the identification result, and the matching confidence is output at the same time.

[0083] S312. The direction of arrival of the vibration characteristic sequence of the structure is estimated by using the inversion algorithm, and the continuous position coordinates of the vibration source in the geographic coordinate system are calculated to form the operation trajectory.

[0084] Specifically, when using the inversion algorithm to estimate the direction of arrival (DOA) of a structural vibration feature sequence, the arrival time difference of the structural vibration signal generated by the same vibration source between different nodes is first calculated based on the geographical coordinates of each detector node. A set of DOA estimation equations is constructed based on the hyperbolic positioning principle. The azimuth and elevation angles of the vibration source relative to the array are inverted by solving a nonlinear least squares problem. Then, combined with the known geographical coordinates of the detector nodes and the vibration propagation velocity model, the DOA information is converted into the spatial position coordinates of the vibration source in the geographical coordinate system. Subsequently, the above calculation process is repeated within a continuous time window to obtain a discrete set of position points of the vibration source that changes over time. Finally, Kalman filtering is used to smooth the discrete set of position points and associate it with the trajectory. The filtered position points are connected sequentially in time order to form a continuous operating trajectory of the vibration source in the geographical coordinate system.

[0085] S313. Based on the repetition frequency and amplitude characteristics of the periodic impact pulses in the structural vibration characteristic sequence, determine the excitation state and cumulative number of compaction passes of the machine as the working state; integrate the above-mentioned machine type, working trajectory and working state into a set of machine working behavior parameters.

[0086] Specifically, firstly, envelope detection and peak detection are performed on the structural vibration characteristic sequence to identify periodic impact pulses and record their occurrence time and amplitude. The repetition frequency is obtained by calculating the time interval between adjacent impact pulses. When the repetition frequency is stable within the preset excitation frequency range (e.g., 8-15Hz) and the pulse amplitude exceeds the preset excitation start amplitude threshold, the machine is determined to be in the excitation start state; otherwise, it is determined to be in the excitation stop state. At the same time, the effectiveness of the compaction operation is judged based on the attenuation characteristics of the pulse amplitude. When the pulse amplitude gradually decreases with the increase of the number of compaction passes, it is determined that a single effective compaction is completed. Based on the cumulative count of impact pulses and the spatial repeatability of the operation trajectory, the cumulative number of compaction passes is accumulated for each effective passage in the same spatial area. Finally, the excitation state and the cumulative number of compaction passes are output together as the machine's operating state parameters.

[0087] In practical applications, after identifying the type of machinery, further analysis of the detailed features of the time-frequency energy peak envelope helps determine the real-time operating status of the machinery. For example:

[0088] For road rollers, the vibration excitation device is activated based on the change in the amplitude of the impact pulse (a sudden increase in amplitude of more than 3 times indicates that the vibration is activated), and the change in the travel speed is determined based on the change in the pulse repetition frequency.

[0089] For pavers, the stability of vibration signals is used to determine whether they are in a continuous operation state.

[0090] For material transport vehicles, the change in energy amplitude determines whether the vehicle is empty or fully loaded, and the presence or absence of pulses determines whether the vehicle is in motion or idling.

[0091] like Figure 4 As shown, in a preferred embodiment of the present invention, the method for reconstructing the high-frequency micro-vibration feature sequence includes the following steps:

[0092] S321. Perform peak detection and pulse interval analysis on the high-frequency micro-vibration feature sequence, extract the time series of gait impact events, and construct a gait event sequence;

[0093] Specifically, firstly, the high-frequency micro-vibration characteristic sequence is bandpass filtered to retain signal components that match the frequency band of human gait. Then, an adaptive thresholding method is used to detect peak values ​​in the filtered signal, marking local maxima exceeding a preset amplitude threshold as candidate gait impact points, and recording the timestamp corresponding to each candidate impact point. Subsequently, pulse interval analysis is performed on the candidate impact points to calculate the time interval between adjacent candidate impact points. When the time interval is within the interval range corresponding to the normal human gait frequency (0.4 seconds to 0.8 seconds) and the coefficient of variation of multiple consecutive intervals is less than a preset stability threshold, these candidate impact points are confirmed as valid gait impact events; otherwise, they are discarded as noise. Finally, the confirmed valid gait impact events are arranged in chronological order, and the occurrence time and pulse amplitude of each event are recorded to form a time series of gait impact events, i.e., a gait event sequence. This gait event sequence reflects the walking rhythm and gait frequency characteristics of construction workers, providing basic data for subsequent gait behavior reconstruction.

[0094] S322. The arrival time difference of multiple points in the high-frequency micro-vibration feature sequence is calculated using the triangulation method, the spatial coordinates corresponding to each gait impact event are calculated, and the movement path is generated.

[0095] Specifically, firstly, when a gait impact event is detected in the high-frequency micro-vibration feature sequence, the arrival times of the signals at at least three geophone nodes for the gait impact event are extracted, and the relative time difference of arrival between each node is calculated. Then, based on the known geographical coordinates of each geophone node and combined with the ground vibration propagation velocity model, a hyperbolic positioning equation system with the time difference of arrival as the observation is established, and the geographical coordinates of the gait impact event are obtained by solving the equation system. Subsequently, the above calculation is performed on the continuously detected gait impact events in sequence to obtain the discrete geographical coordinate point set corresponding to each gait impact event. Finally, according to the occurrence time sequence of the gait impact events, the discrete geographical coordinate points are connected sequentially, and Kalman filtering is used to smooth the connected trajectory, eliminating outliers caused by positioning errors, and generating a continuous movement path for the construction personnel.

[0096] The expression for the ground vibration propagation velocity model is as follows:

[0097] ;

[0098] In the formula, v is the actual propagation speed of the ground vibration signal in the construction site; v0 is the reference wave velocity, which is calculated by selecting a calibration point in advance at the construction site, conducting a hammer test, and measuring the distance and time difference between the hammer impact point and the detector node. The soil moisture correction coefficient is determined by acquiring real-time moisture values ​​through soil moisture sensors deployed at the construction site and based on a pre-calibrated moisture-wave velocity relationship curve. Soil volumetric moisture content is collected in real time by a soil moisture sensor; d is the distance attenuation factor, which is preset based on the geological survey report of the construction site; d is the propagation distance between the vibration source and the detector node.

[0099] The hyperbolic positioning equations include the hyperbolic positioning equations corresponding to each detector node; for the i-th and j-th detector nodes, the following hyperbolic positioning equations are established:

[0100] ;

[0101] In the formula, (x,y) are the geographic coordinates of the gait collision event to be solved; (x i ,y i (x) represents the known geographic coordinates of the i-th detector node; j ,y j ) represents the known geographic coordinates of the j-th detector node; t i The time when the gait impact event signal arrives at the i-th detector node is recorded by the high-precision clock of the detector node; t j The time when the gait impact event signal arrives at the j-th detector node is recorded by the high-precision clock of the detector node. In this embodiment of the invention, for at least three detector nodes, the above hyperbolic positioning equations are combined to form a set of equations, which are then solved using the least squares method to obtain the geographical coordinates of the gait impact event.

[0102] S323. Based on the consistency of the interval duration between adjacent gait impact events in the gait event sequence, calculate gait rhythm features to identify the person's identity and abnormal gait; integrate the above gait event sequence, movement path, and gait rhythm features into a gait behavior parameter set;

[0103] Specifically, firstly, timestamps of adjacent gait impact events are extracted from the gait event sequence, and the interval duration between each adjacent gait impact event is calculated to form an interval duration sequence. Then, statistical analysis is performed on the interval duration sequence to calculate the mean, standard deviation, and coefficient of variation of the interval duration, and the temporal variation trend and frequency domain distribution characteristics of the interval duration are extracted to jointly constitute gait rhythm features. In terms of identity recognition, the gait rhythm features of multiple consecutive stride cycles of the current construction worker are matched with a pre-established personnel gait feature database, and a dynamic time warping algorithm is used to calculate the similarity. When the similarity exceeds a preset identity recognition threshold, the corresponding personnel identity is identified. In terms of abnormal gait recognition, when multiple consecutive interval durations in the interval duration sequence deviate from the preset normal gait range by more than a preset deviation threshold, or when the coefficient of variation of the interval duration sequence exceeds a preset stability threshold, it is determined to be an abnormal gait state. Combined with the elevation change rate of the movement path at the corresponding time, it is further determined whether an abnormal event such as a fall or stagger has occurred, triggering an abnormal gait fall risk alarm.

[0104] like Figure 5 As shown, in a preferred embodiment of the present invention, the step of synchronously mapping the set of mechanical operation behavior parameters and the set of gait behavior parameters to a unified three-dimensional geographic information spatiotemporal grid to generate a dynamic construction activity coupling field, namely step S400, specifically includes:

[0105] S410. Construct a three-dimensional geographic information spatiotemporal grid covering the entire construction site;

[0106] Specifically, firstly, a digital topographic map and a construction site layout plan are acquired. The construction site is then divided into planar sections with a preset grid spacing (e.g., 1 meter × 1 meter). Elevation data from the digital topographic map is used to assign corresponding elevation information to each planar grid cell, forming a three-dimensional geographic information grid base. Next, a data structure is established for each grid cell, including geographic coordinate fields (X and Y coordinate values), elevation information fields (Z coordinate values), and a temporal status identifier field. The temporal status identifier is further subdivided into four subfields: last update timestamp, current occupancy status, mechanical activity intensity, and personnel activity intensity. Initially, all fields are set to empty or zero. Finally, during operation, a timestamp alignment mechanism is used to synchronously update the temporal status identifier field of the grid cell whenever a set of mechanical operation behavior parameters or gait behavior parameters is mapped to it. This records the latest update time and activity status, forming a spatiotemporal grid model that dynamically evolves over time, providing a unified spatiotemporal reference and status record carrier for the subsequent construction of the construction activity coupling field.

[0107] S420. Map the operation trajectory and operation status of the mechanical operation behavior parameter set to the three-dimensional geographic information spatiotemporal grid to generate a dynamically updated mechanical operation heat map.

[0108] Specifically, firstly, based on the work trajectory coordinates in the set of mechanical operation behavior parameters, the grid cell where the machine is located at the current moment is determined, and the geographical coordinates of the grid cell are associated with the real-time location of the machine. Then, the machine type and operation status in the set of mechanical operation behavior parameters are extracted, the machine type is encoded as a visual color index, and the excitation status and cumulative value of compaction passes in the operation status are converted into thermal intensity weight values, which are written into the temporal status identifier field of the corresponding grid cell, and the mechanical activity intensity and last update timestamp of the grid cell are updated. Subsequently, for continuous grid cells traversed by the work trajectory, the Gaussian kernel density estimation method is used to perform spatial smoothing interpolation on the thermal intensity of adjacent grid cells to form a spatial continuous distribution of mechanical activity intensity. Finally, based on the mechanical activity intensity value of each grid cell, a dynamically updated mechanical operation heat map is generated in real time according to a preset color level mapping rule (such as low intensity corresponding to blue, medium intensity corresponding to yellow, and high intensity corresponding to red), and the intensity of grid cells that have not been updated for a long time is attenuated over time to ensure that the heat map reflects the latest spatiotemporal distribution of mechanical operations.

[0109] S430. Map the movement paths and gait event sequences in the gait behavior parameter set to the three-dimensional geographic information spatiotemporal grid to generate a dynamically updated personnel distribution density field.

[0110] Specifically, firstly, based on the movement path coordinates in the gait behavior parameter set, the grid cell where the construction worker is located at each moment is determined, and the geographical coordinates of the grid cell are associated with the real-time location of the worker. Then, the gait event sequence in the gait behavior parameter set is extracted, and the activity intensity weight of the worker in each grid cell is calculated based on the time density and pulse amplitude of the gait impact event. The number of workers and the activity intensity are written into the temporal status identifier field of the corresponding grid cell, and the activity intensity and last update timestamp of the worker in the grid cell are updated. For the case where there are multiple workers in the same grid cell, the number of workers and the activity intensity are superimposed in an accumulative manner. Subsequently, a kernel density estimation algorithm is used to perform spatial smoothing interpolation on the activity intensity of workers in adjacent grid cells, expanding the discrete worker location points into a continuous spatial distribution, forming a spatial continuous density field of worker distribution. Finally, based on the worker density value of each grid cell, a dynamically updated worker distribution density field is generated in real time according to a preset color level mapping rule, and the density of grid cells that have not been updated for a long time is subjected to density decay processing over time to ensure that the worker distribution density field reflects the latest spatiotemporal distribution status of the construction workers.

[0111] S440. By using timestamp alignment and spatial interpolation algorithms, the mechanical operation heat map and the personnel distribution density field are fused to generate a dynamic construction activity coupling field.

[0112] Specifically, firstly, the last update timestamp of the mechanical activity intensity and personnel activity intensity of each grid cell in the mechanical operation heatmap and personnel distribution density field are extracted. For grid cells with timestamp differences exceeding a preset synchronization threshold, linear interpolation is used to extrapolate the earlier data to ensure that the two types of data involved in the fusion are synchronized in the time dimension. Then, for the same grid cell after timestamp synchronization, the mechanical activity intensity and personnel activity intensity are weighted and superimposed, and the grid cell is marked as a human-machine interaction area, a mechanical independent operation area, a personnel independent activity area, or an empty area based on the superposition result. Subsequently, for the grid cells marked as human-machine interaction areas, the spatial boundary of the dynamic restricted area is calculated by combining the mechanical type and operation status. Spatial interpolation enhancement processing is performed on the personnel activity intensity near the boundary of the dynamic restricted area to highlight the human-machine proximity risk. Finally, the fused grid cells are visualized and rendered according to the interaction type and risk level to generate a dynamic construction activity coupling field that includes human-machine spatial relationship, interaction intensity, and dynamic restricted area boundary, providing a unified spatiotemporal coupling data foundation for the compliance judgment of the subsequent safety behavior rule engine.

[0113] In another embodiment of the present invention, the safety behavior rules include:

[0114] Intrusion detection rules for restricted areas: When the geographic coordinates of any person’s gait behavior parameter set fall into a grid cell marked as a dynamic restricted area in the mechanical operation behavior parameter set, an intrusion alarm is triggered; the dynamic restricted area can be determined in real time according to the type of machinery and the operation status.

[0115] Human-machine interaction distance violation rules: Calculate the Euclidean distance between the center of mass of the person in the gait behavior parameter set and the center of mass of the machine in the machine operation behavior parameter set. When the Euclidean distance is less than the preset safety threshold, trigger a close-range operation violation alarm.

[0116] like Figure 6 As shown, in another embodiment of the present invention, an assessment system for compliance of safety behaviors at highway construction sites is also provided, used to implement the assessment method for compliance of safety behaviors at highway construction sites as described above, specifically including:

[0117] The vibration data acquisition module 10 is used to acquire ground coupled vibration data in real time; the ground coupled vibration data includes ground vibration signals collected in real time by multiple geophone nodes deployed at the construction site.

[0118] The feature extraction module 20 is used to remove interference features from the ground coupled vibration data and extract the structural vibration feature sequence induced by mechanical operation and the high-frequency micro-vibration feature sequence induced by personnel gait.

[0119] The feature reconstruction module 30 is used to reconstruct the structural vibration feature sequence and the high-frequency micro-vibration feature sequence respectively to obtain a mechanical operation behavior parameter set and a gait behavior parameter set; the mechanical operation behavior parameter set includes mechanical type, operation trajectory and operation status; the gait behavior parameter set includes gait event sequence, movement path and gait rhythm features;

[0120] The coupling field construction module 40 is used to synchronously map the mechanical operation behavior parameter set and the gait behavior parameter set to a unified three-dimensional geographic information spatiotemporal grid to generate a dynamic construction activity coupling field.

[0121] The compliance assessment module 50 is used to determine the human-machine interaction distance and the risk of intrusion into the restricted area of ​​mechanical operation in real time in the construction activity coupling field based on preset safety behavior rules, and obtain the safety behavior compliance assessment results.

[0122] like Figure 7 As shown, in a preferred embodiment of the present invention, the feature extraction module 20 specifically includes:

[0123] The vehicle interference removal unit 21 is used to remove non-construction vehicle interference from the ground coupled vibration data based on the beamforming algorithm to obtain the first processed data;

[0124] The environmental interference identification unit 22 is used to identify environmental interference based on the energy distribution entropy method, and to adaptively adjust the parameters of the first processed data to obtain the second processed data.

[0125] The feature separation unit 23 is used to separate the structural vibration feature sequence and the high-frequency micro-vibration feature sequence from the second processed data based on the variational mode decomposition method.

[0126] like Figure 8 As shown, in a preferred embodiment of the present invention, the feature reconstruction module 30 specifically includes:

[0127] The mechanical type identification unit 311 is used to extract the time-frequency energy peak envelope in the structural vibration feature sequence to identify the mechanical type.

[0128] The operation trajectory generation unit 312 is used to estimate the direction of arrival of the vibration feature sequence of the structure using an inversion algorithm, calculate the continuous position coordinates of the vibration source in the geographic coordinate system, and form an operation trajectory.

[0129] The operation status determination unit 313 is used to determine the excitation state and cumulative number of rolling passes of the machine based on the repetition frequency and amplitude characteristics of the periodic impact pulses in the structural vibration characteristic sequence, and to use this as the operation status.

[0130] The event sequence construction unit 321 is used to perform peak detection and pulse interval analysis on the high-frequency micro-vibration feature sequence, extract the time sequence of gait impact events, and construct a gait event sequence.

[0131] The movement path generation unit 322 is used to calculate the time difference of arrival of multiple points in the high-frequency micro-vibration feature sequence using the triangulation method, solve the spatial coordinates corresponding to each gait impact event, and generate a movement path.

[0132] The rhythm feature calculation unit 323 is used to calculate gait rhythm features based on the consistency of the interval duration between adjacent gait impact events in the gait event sequence, and to identify the person's identity and abnormal gait.

[0133] like Figure 9 As shown, in a preferred embodiment of the present invention, the coupling field construction module 40 specifically includes:

[0134] Spatiotemporal grid construction unit 41 is used to construct a three-dimensional geographic information spatiotemporal grid covering the entire construction site;

[0135] The first mapping unit 42 is used to map the operation trajectory and operation status of the mechanical operation behavior parameter set to the three-dimensional geographic information spatiotemporal grid to generate a dynamically updated mechanical operation heat map.

[0136] The second mapping unit 43 is used to map the movement paths and gait event sequences in the gait behavior parameter set to the three-dimensional geographic information spatiotemporal grid to generate a dynamically updated personnel distribution density field.

[0137] The human-machine fusion unit 44 is used to fuse the mechanical operation heat map with the personnel distribution density field through timestamp alignment and spatial interpolation algorithm to generate a dynamic construction activity coupling field.

[0138] It should be noted that the above modules and units can be implemented as a computer program, which can run on a computer device. The computer device's memory can store the computer program that makes up the modules or units, enabling the processor to execute the various steps of the above method.

[0139] It should be understood that although the steps in the flowcharts of the embodiments of the present invention are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in each embodiment may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.

[0140] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods.

[0141] The above embodiments merely illustrate several implementation methods of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this patent should be determined by the appended claims.

Claims

1. A method for assessing the compliance of safety behaviors at highway construction sites, characterized in that, Includes the following steps: Real-time acquisition of ground coupled vibration data; the ground coupled vibration data includes ground vibration signals collected in real time by multiple geophone nodes deployed at the construction site; Interference features are removed from the ground coupled vibration data, and the structural vibration feature sequence induced by mechanical operation and the high-frequency micro-vibration feature sequence induced by human gait are extracted. The structural vibration feature sequence and the high-frequency micro-vibration feature sequence are reconstructed to obtain a mechanical operation behavior parameter set and a gait behavior parameter set. The mechanical operation behavior parameter set includes the machine type, operation trajectory, and operation status. The gait behavior parameter set includes the gait event sequence, movement path, and gait rhythm features. The mechanical operation behavior parameter set and the gait behavior parameter set are synchronously mapped to a unified three-dimensional geographic information spatiotemporal grid to generate a dynamic construction activity coupling field. Based on preset safety behavior rules, the human-machine interaction distance and the risk of intrusion into the restricted area of ​​mechanical operation are determined in real time in the construction activity coupling field, and the safety behavior compliance assessment results are obtained.

2. The method for assessing compliance with safety behaviors at highway construction sites according to claim 1, characterized in that, The steps of removing interference features from the ground coupled vibration data and extracting the structural vibration feature sequence induced by mechanical operations and the high-frequency micro-vibration feature sequence induced by human gait specifically include: Based on the beamforming algorithm, non-construction vehicle interference is removed from the ground coupled vibration data to obtain the first processed data; Based on the energy distribution entropy method, environmental interference is identified, and adaptive parameter adjustments are made to the first processed data to obtain the second processed data. Based on variational mode decomposition, structural vibration characteristic sequences and high-frequency micro-vibration characteristic sequences are separated from the second processed data.

3. The method for assessing compliance with safety behaviors at highway construction sites according to claim 1, characterized in that, The method for reconstructing the structural vibration characteristic sequence includes the following steps: Extract the time-frequency energy peak envelope from the structural vibration feature sequence to identify the mechanical type; The direction of arrival of the vibration characteristic sequence of the structure is estimated using an inversion algorithm, and the continuous position coordinates of the vibration source in the geographic coordinate system are calculated to form the operation trajectory. Based on the repetition frequency and amplitude characteristics of the periodic impact pulses in the structural vibration characteristic sequence, the excitation state and cumulative number of compaction passes of the machinery are determined as the working state.

4. The method for assessing compliance with safety behaviors at highway construction sites according to claim 1, characterized in that, The method for reconstructing the high-frequency micro-vibration feature sequence includes the following steps: Peak detection and pulse interval analysis are performed on the high-frequency micro-vibration feature sequence to extract the time series of gait impact events and construct a gait event sequence; The arrival time difference of multiple points in the high-frequency micro-vibration feature sequence is calculated using the triangulation method, and the spatial coordinates corresponding to each gait impact event are solved to generate a movement path; Based on the consistency of the interval duration between adjacent gait impact events in the gait event sequence, gait rhythm features are calculated to identify personnel identity and abnormal gait.

5. The method for assessing compliance with safety behaviors at highway construction sites according to claim 1, characterized in that, The steps of synchronously mapping the set of mechanical operation behavior parameters and the set of gait behavior parameters onto a unified three-dimensional geographic information spatiotemporal grid to generate a dynamic construction activity coupling field specifically include: Construct a three-dimensional geographic information spatiotemporal grid covering the entire construction site; The operation trajectory and operation status of the mechanical operation behavior parameter set are mapped to the three-dimensional geographic information spatiotemporal grid to generate a dynamically updated mechanical operation heat map. The movement paths and gait event sequences in the gait behavior parameter set are mapped to the three-dimensional geographic information spatiotemporal grid to generate a dynamically updated personnel distribution density field. By using timestamp alignment and spatial interpolation algorithms, the mechanical operation heatmap and the personnel distribution density field are fused to generate a dynamic construction activity coupling field.

6. A system for assessing compliance with safety behaviors at highway construction sites, used to implement the assessment method for assessing compliance with safety behaviors at highway construction sites as described in any one of claims 1-5, characterized in that, include: The vibration data acquisition module is used to acquire ground coupled vibration data in real time; the ground coupled vibration data includes ground vibration signals collected in real time by multiple geophone nodes deployed at the construction site. The feature extraction module is used to remove interfering features from the ground coupled vibration data and extract the structural vibration feature sequence induced by mechanical operation and the high-frequency micro-vibration feature sequence induced by personnel gait. The feature reconstruction module is used to reconstruct the structural vibration feature sequence and the high-frequency micro-vibration feature sequence respectively to obtain a mechanical operation behavior parameter set and a gait behavior parameter set; the mechanical operation behavior parameter set includes the machine type, operation trajectory and operation status; the gait behavior parameter set includes the gait event sequence, movement path and gait rhythm features; The coupling field construction module is used to synchronously map the set of mechanical operation behavior parameters and the set of gait behavior parameters to a unified three-dimensional geographic information spatiotemporal grid to generate a dynamic construction activity coupling field. The compliance assessment module is used to determine the human-machine interaction distance and the risk of intrusion into the restricted area of ​​mechanical operation in real time in the construction activity coupling field based on preset safety behavior rules, and obtain the safety behavior compliance assessment results.

7. The assessment system for compliance of safety behaviors at highway construction sites according to claim 6, characterized in that, The feature extraction module specifically includes: The vehicle interference removal unit is used to remove non-construction vehicle interference from the ground coupled vibration data based on the beamforming algorithm to obtain the first processed data; An environmental interference identification unit is used to identify environmental interference based on the energy distribution entropy method, and to adaptively adjust the parameters of the first processed data to obtain the second processed data. The feature separation unit is used to separate the structural vibration feature sequence and the high-frequency micro-vibration feature sequence from the second processed data based on the variational mode decomposition method.

8. The assessment system for compliance of safety behaviors at highway construction sites according to claim 6, characterized in that, The feature reconstruction module specifically includes: The mechanical type identification unit is used to extract the time-frequency energy peak envelope from the structural vibration feature sequence to identify the mechanical type. The operation trajectory generation unit is used to estimate the direction of arrival of the vibration feature sequence of the structure using an inversion algorithm, calculate the continuous position coordinates of the vibration source in the geographic coordinate system, and form the operation trajectory. The operation status determination unit is used to determine the excitation state and cumulative number of compaction passes of the machine based on the repetition frequency and amplitude characteristics of the periodic impact pulses in the structural vibration characteristic sequence, and to use this as the operation status.

9. The assessment system for compliance of safety behaviors at highway construction sites according to claim 6, characterized in that, The feature reconstruction module specifically includes: The event sequence construction unit is used to perform peak detection and pulse interval analysis on the high-frequency micro-vibration feature sequence, extract the time series of gait impact events, and construct the gait event sequence. The movement path generation unit is used to calculate the time difference of arrival of multiple points in the high-frequency micro-vibration feature sequence using the triangulation method, solve the spatial coordinates corresponding to each gait impact event, and generate a movement path. The rhythm feature calculation unit is used to calculate gait rhythm features based on the consistency of the interval duration between adjacent gait impact events in the gait event sequence, and to identify personnel identity and abnormal gait.

10. The assessment system for compliance of safety behaviors at highway construction sites according to claim 6, characterized in that, The coupling field construction module specifically includes: Spatiotemporal grid construction unit, used to construct a three-dimensional geographic information spatiotemporal grid covering the entire construction site; The first mapping unit is used to map the operation trajectory and operation status of the mechanical operation behavior parameter set to the three-dimensional geographic information spatiotemporal grid to generate a dynamically updated mechanical operation heat map. The second mapping unit is used to map the movement paths and gait event sequences in the gait behavior parameter set to the three-dimensional geographic information spatiotemporal grid, thereby generating a dynamically updated personnel distribution density field. The human-machine fusion unit is used to fuse the mechanical operation heat map with the personnel distribution density field through timestamp alignment and spatial interpolation algorithms to generate a dynamic construction activity coupling field.

Images