Multi-scene aerial work personnel positioning and trajectory data fusion analysis system

Abstract

The present application relates to high-precision positioning technology field, specifically for multi-scene high-altitude operation personnel positioning and trajectory data fusion analysis system, the system includes phase acquisition module, angle reconstruction module, trajectory generation module, main shaft weight module and trajectory analysis module.In the present application, the cooperative calibration relationship is established by the carrier phase difference change and the air pressure height change, the abnormal deviation is dynamically corrected, the time sequence data is continuously stable, the path pointing information is formed by introducing the structure boundary constraint and the incident direction inference, the space propagation path has the distinguishable characteristics, the path confusion caused by the environmental interference is reduced, the missing nodes are supplemented according to the height change and the time correlation, the interrupted fragment is connected, the continuous three-dimensional trajectory form is constructed, the direction weight is distributed by the position difference and the error amplitude, and the change trend is matched, so that the trajectory state has the ordered correlation characteristics, the consistency of the trajectory change and the behavior discrimination ability are enhanced.

Description

Technical Field

[0001] This invention relates to the field of high-precision positioning technology, and in particular to a multi-scenario high-altitude worker positioning and trajectory data fusion analysis system. Background Technology

[0002] The field of high-precision positioning technology involves determining the target location through spatial signal measurement. Specifically, this includes calculating coordinates based on the time difference of satellite signal propagation, determining distance based on base station signal strength, obtaining acceleration and integrating displacement using an inertial measurement unit, and performing position fusion processing using multi-source sensor data. This technology covers location acquisition, time synchronization, coordinate calculation, and trajectory recording, and is widely used in scenarios such as personnel positioning, equipment tracking, and path analysis. Among these, the traditional multi-scenario high-altitude work personnel positioning and trajectory data fusion analysis system refers to receiving satellite navigation signals and obtaining latitude and longitude coordinates in a high-altitude work environment using a worn terminal, combining this with ground base station signal strength for assisted positioning, continuously recording multiple coordinate points in a time series to form trajectory data, stitching and storing the trajectory data according to preset time intervals, using weighted average calculations of multiple sets of coordinate data to correct single measurement errors, and aligning positioning data from different sources using timestamps to form a complete record of personnel movement trajectory data.

[0003] Trajectories formed by splicing single coordinate records with time series lack continuity constraints in cases of signal anomalies or short-term missing data, making it prone to trajectory fragment breaks or disordered connections. Furthermore, error correction relies on overall equalization processing, which is insufficient to reflect differences in directional changes. In scenarios with complex spatial structures or frequent height changes, trajectory morphology may be distorted or offset. Additionally, multi-source data is fused solely based on time alignment, lacking constraints on the correlation between changing trends and spatial relationships, resulting in inconsistencies in the expression of behavioral trajectories. For example, skipped connections or trajectory deviations may occur in continuous operation paths, affecting the accuracy of subsequent analysis. Summary of the Invention

[0004] To address the technical problems existing in the prior art, embodiments of the present invention provide a multi-scenario high-altitude worker positioning and trajectory data fusion analysis system.

[0005] On the one hand, a multi-scenario high-altitude worker positioning and trajectory data fusion analysis system is provided, which includes: The phase acquisition module acquires the carrier phase difference sequence and barometer height value sequence received by the terminal worn by the high-altitude worker, extracts the carrier phase difference change and height value change at adjacent moments, replaces the deviation value according to the offset, and obtains a continuous phase sequence. Based on the phase continuity sequence, the angle reconstruction module obtains the incident angle and propagation time at the end of the tower crane boom and the column structure of the lifting platform, compares it with the structural boundary angle range and associates it with the propagation direction to obtain the path direction sequence. Based on the path direction sequence, the trajectory generation module obtains the time node positioning coordinate sequence and the height change sequence, analyzes the connection status and supplements the interruption position matching height change sequence to obtain the three-dimensional trajectory structure sequence. Based on the three-dimensional trajectory structure sequence, the main axis weighting module extracts the multi-source trajectory coordinates and position differences at time nodes, compares them with the error amplitude threshold, assigns directional weights and associates them with directions to obtain the trajectory directional weight sequence. Based on the trajectory direction weight sequence, the trajectory analysis module extracts the trajectory change status at time nodes, compares the change trends of adjacent nodes, and associates discontinuous nodes to match the time sequence, thereby obtaining the trajectory behavior association structure.

[0006] As a further embodiment of the present invention, the phase continuity sequence includes a set of phase continuity values, a jump correction marker, a time index label, a height constraint label, and an offset correction parameter; the path direction sequence includes a set of direction vectors, an incident angle identifier, a propagation time label, a direction consistency marker, and a boundary angle constraint; the three-dimensional trajectory structure sequence includes a set of spatial coordinate points, a timestamp sequence, a height level identifier, trajectory connection relationships, and completion node markers; the trajectory direction weight sequence includes direction weight values, error amplitude labels, direction category labels, continuity scores, and weight distribution parameters; and the trajectory behavior association structure includes behavior state labels, a change trend index, discontinuous node markers, a rearranged time chain, and association relationships.

[0007] As a further aspect of the present invention, the replacement deviation value in the comparison deviation situation refers to replacing the detected abnormal deviation data with the re-acquired and corrected data by comparing the degree of data deviation; The reference structural boundary angle range refers to comparing the measured incident angle with the preset angle range of the preset structure and eliminating invalid angle data that exceeds the structural constraints.

[0008] As a further aspect of the present invention, the trend of adjacent node changes refers to the continuous analysis of the direction and magnitude of changes of adjacent nodes in the time series; The discontinuous node refers to a trajectory node that has abrupt changes and disrupts continuity with the preceding and following nodes in terms of temporal and spatial variations.

[0009] As a further aspect of the present invention, the phase acquisition module includes: The phase change extraction submodule acquires the carrier phase difference sequence and barometer height value sequence received by the terminal worn by the high-altitude worker, extracts the carrier phase difference change and height value change at adjacent moments, compares the differences between adjacent data in the carrier phase difference sequence, compares the differences between adjacent data in the barometer height value sequence, and obtains the phase change amplitude and height change amplitude. The interval identification submodule, based on the phase change amplitude and the height change amplitude, compares the phase change amplitude with a preset phase jump threshold to identify the deviation interval, and compares the height change amplitude with a preset height change threshold to identify the smooth interval, thus obtaining the phase deviation interval index and the height smooth interval index. The phase correction submodule, based on the phase deviation interval index and the height smooth interval index, uses an interpolation fitting algorithm to reconstruct the corresponding deviation interval position in the carrier phase difference sequence, and replaces the repeated acquisition data with the original data after point-by-point offset comparison, to obtain a continuous phase sequence.

[0010] As a further aspect of the present invention, the angle reconstruction module includes: The incident angle extraction submodule, based on the phase continuity sequence, obtains the incident angle and propagation time of the multi-frequency signal at the end of the tower crane boom and the column structure of the lifting operation platform, extracts the incident angle change, and compares the incident angle data with the angle range of the structural boundary point by point, removes the incident angle data that exceeds the structural boundary range, and obtains the angle change amplitude sequence. The direction association submodule maps the incident angle data at time nodes in the angle change amplitude sequence to the corresponding propagation time direction based on the angle change amplitude sequence, and determines the direction pointing change according to the time sequence correspondence to obtain the direction pointing sequence. The consistency determination submodule, based on the direction pointing sequence, compares the direction pointing of adjacent time nodes point by point, identifies the direction change offset segment by comparing the range of direction angle change with the preset direction offset threshold, and corrects the direction to obtain the path direction sequence.

[0011] As a further aspect of the present invention, the trajectory generation module includes: Based on the path direction sequence, the trajectory sorting submodule obtains the corresponding time node positioning coordinate sequence and height change sequence, sorts the positioning coordinate sequence according to the timestamp order and extracts the trajectory point sequence, analyzes the temporal continuity of the trajectory point sequence, and obtains the trajectory point time sequence. The connection determination submodule identifies the connection status based on the trajectory point time sequence, compares the spatial position and time interval of adjacent trajectory points, and obtains the trajectory interruption position sequence according to the spatial interval threshold and the time interval threshold. The trajectory matching submodule interpolates and supplements the trajectory interruption position sequence and compares it point by point with the height change sequence. The spatial changes of the supplemented trajectory points are consistent with the original trajectory points, thus obtaining a three-dimensional trajectory structure sequence.

[0012] As a further aspect of the present invention, the spindle weighting module includes: The difference extraction submodule obtains the coordinates of multiple sources at the same time node based on the three-dimensional trajectory structure sequence, compares the spatial position of the multi-source trajectory coordinates point by point, extracts the directional position difference according to the coordinate axis difference, and matches them according to the time node order to obtain the position difference sequence. The direction recognition submodule extracts the magnitude of the direction position difference based on the position difference sequence, compares the magnitude data with the error magnitude threshold item by item, classifies the direction category according to the range of magnitude change, and obtains the direction category sequence corresponding to the time node. The weight allocation submodule assigns weights to the direction category based on the direction category sequence, associates and maps the direction relationships, compares the continuity of the direction relationships between adjacent time nodes, identifies the offset direction based on the range of direction changes, and obtains the trajectory direction weight sequence.

[0013] As a further aspect of the present invention, the trajectory analysis module includes: The state extraction submodule extracts the trajectory change state at time nodes based on the trajectory direction weight sequence, and maps the direction weights of time nodes in the trajectory direction weight sequence to the corresponding trajectory position changes to obtain the trajectory change sequence. The trend comparison submodule, based on the trajectory change sequence, compares the trajectory changes at adjacent time nodes, and compares the range of directional changes with a preset change benchmark to identify discontinuous change nodes and obtain a set of change nodes. The association matching submodule, based on the set of changing nodes, associates discontinuous time nodes, reorganizes the time order of nodes in the set of changing nodes, and compares the overall trajectory change continuity to obtain the trajectory behavior association structure.

[0014] As a further aspect of the present invention, in the process of corresponding the time node direction weights and the corresponding trajectory position changes: the time interval between adjacent time nodes is limited to a preset time window for matching one by one, and a sliding comparison is performed on three consecutive time nodes in the trajectory direction weight sequence. In the process of reorganizing the time order of nodes in the set of changed nodes: the nodes are sorted from earliest to latest according to the timestamp, and the time interval between adjacent time nodes after sorting is checked for continuity. Time nodes whose time interval exceeds the preset time window are removed.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, a collaborative verification relationship is established by changing the carrier phase difference and the air pressure altitude to dynamically correct abnormal offsets, maintain the continuity and stability of time series data, and introduce structural boundary constraints and incident direction inference to form path pointing information, so that the propagation path in space has discriminative characteristics, reducing path confusion caused by environmental interference. Missing nodes are supplemented based on altitude changes and time correlation, the connection of interrupted segments is improved, and a continuous three-dimensional trajectory shape is constructed. At the same time, the directional weights are assigned by position differences and error amplitudes and matched with the changing trend, so that the trajectory state has orderly correlation characteristics, enhancing the consistency of trajectory changes and the ability to discriminate behavior. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a system flowchart of the present invention; Figure 2 This is a system block diagram of the present invention; Figure 3 This is a flowchart of the phase acquisition module in this invention; Figure 4 This is a flowchart of the angle reconstruction module in this invention; Figure 5 This is a flowchart of the trajectory generation module in this invention; Figure 6 This is a flowchart of the spindle weighting module in this invention; Figure 7 This is a flowchart of the trajectory analysis module in this invention. Detailed Implementation

[0018] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0019] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0020] This invention provides a multi-scenario high-altitude worker positioning and trajectory data fusion analysis system, such as... Figure 1-2 The diagram shown illustrates a multi-scenario high-altitude worker positioning and trajectory data fusion analysis system. This system includes: The phase acquisition module acquires the carrier phase difference sequence and barometer altitude value sequence received by the terminal worn by the high-altitude worker, extracts the carrier phase difference change at adjacent time moments and outputs the phase change sequence, compares the phase change sequence with the phase jump threshold point by point, identifies continuous deviation intervals, extracts altitude value changes and outputs the altitude change sequence and identifies smooth intervals, repeatedly acquires the deviation phase in the smooth intervals, and replaces the deviation value after comparing the deviation situation to obtain a continuous phase sequence. The angle reconstruction module, based on the phase continuous sequence, obtains the incident angle and propagation time of multi-frequency signals at the end of the tower crane boom and the column structure of the lifting platform, extracts the change of incident angle and compares it with the angle range of the structural boundary, identifies the angle data and associates it with the direction of propagation time, analyzes the consistency of the direction change of adjacent time nodes and identifies the deviation direction, and obtains the path direction sequence. The trajectory generation module obtains the positioning coordinate sequence and height change sequence of the corresponding time node based on the path direction sequence, extracts the trajectory point sequence in time order, analyzes the connection status of adjacent trajectory points and identifies the interruption position, supplements the interruption position and matches it with the corresponding height change sequence, compares the consistency of continuous trajectory point changes and identifies deviations from the trajectory segment, and obtains the three-dimensional trajectory structure sequence. The main axis weighting module is based on the three-dimensional trajectory structure sequence. It extracts the coordinates of multi-source trajectories at the same time node, extracts the positional differences between trajectory coordinates, identifies the positional differences sequentially and extracts the change amplitude in each direction, compares the change amplitude with the error amplitude threshold and identifies the direction category, assigns directional relationship weights and associates directions, and identifies the deviation direction based on the continuity of directional relationship between adjacent time nodes to obtain the trajectory direction weight sequence. The trajectory analysis module extracts the trajectory change sequence based on the trajectory direction weight sequence, obtains the trajectory change sequence by extracting the trajectory change state at time nodes, compares the change trends of adjacent nodes in the trajectory change sequence and identifies nodes with discontinuous changes, associates the nodes with discontinuous changes and re-matches the time sequence, analyzes the continuity of trajectory changes, and obtains the trajectory behavior association structure.

[0021] The phase continuity sequence includes a set of phase continuity values, jump correction markers, time index labels, height constraint labels, and offset correction parameters. The path direction sequence includes a set of direction vectors, incident angle identifiers, propagation time labels, direction consistency markers, and boundary angle constraints. The 3D trajectory structure sequence includes a set of spatial coordinate points, a timestamp sequence, height level identifiers, trajectory connection relationships, and completion node markers. The trajectory direction weight sequence includes direction weight values, error amplitude labels, direction category labels, continuity scores, and weight distribution parameters. The trajectory behavior association structure includes behavior state labels, change trend indexes, discontinuous node markers, rearranged time chains, and association relationships.

[0022] Specifically, such as Figure 2 , 3 As shown, the phase acquisition module includes: The phase change extraction submodule acquires the carrier phase difference sequence and barometer height value sequence received by the terminal worn by the high-altitude worker, extracts the carrier phase difference change and height value change at adjacent moments, compares the differences between adjacent data in the carrier phase difference sequence, compares the differences between adjacent data in the barometer height value sequence, and obtains the phase change amplitude and height change amplitude. First, a high-frequency sampling chip extracts the carrier phase difference change between two adjacent sampling times at a period of 50 milliseconds. This sampling frequency is used to obtain high-precision displacement vector observations, providing underlying data support for high-precision positioning. Specifically, by subtracting the current carrier phase observation value of 25.68 cycles from the previous observation value of 25.24 cycles, the phase change amplitude within this time step is found to be 0.44 cycles. Simultaneously, the altitude value sequence output by the barometer is extracted, obtaining the altitude value of 92.45 meters at the current sampling point and the altitude value of 92.41 meters at the adjacent previous sampling point. After subtraction, the altitude change amplitude is found to be 0.04 meters. By traversing the entire data buffer, the numerical differences between adjacent data items in the carrier phase difference sequence are compared. For example, if the phase difference between the 102nd and 103rd points in the sequence jumps from 28.12 cycles to 31.55 cycles, the increment of the difference is calculated. Similarly, by comparing the fluctuation differences between adjacent data in the barometer altitude value sequence, the noise level of the barometer is assessed by analyzing the dispersion of altitude values ​​over a short period of time, and linear compensation is performed on the barometer observations based on the vertical deviation constraint of high-precision positioning. The phase change amplitude and altitude change amplitude are obtained by differential extraction and fluctuation comparison of the dual-source sequences.

[0023] The interval identification submodule identifies deviation intervals by comparing the phase change amplitude with the height change amplitude and the preset phase jump threshold, and identifies smooth intervals by comparing the height change amplitude with the preset height change threshold, thus obtaining the phase deviation interval index and the height smooth interval index. First, a preset phase jump threshold, determined based on electromagnetic multipath interference experiments at the construction site, is retrieved and set to 1.85 cycles. The extracted phase change amplitude of 3.42 cycles is compared with the preset phase jump threshold of 1.85 cycles. The current amplitude value is determined to be greater than the preset threshold, thus identifying this time period as a phase deviation interval. Next, a preset height change threshold is retrieved. This threshold, referencing the vertical displacement range of a person working on a high-altitude platform, is set to 0.25 meters. This threshold serves as the motion judgment benchmark for high-precision positioning along the vertical axis. The real-time height change amplitude of 0.08 meters is compared with the preset height change threshold of 0.25 meters. Since the height change amplitude is less than the preset height change threshold, the person is in a flat state in the vertical dimension, thus identifying this trajectory segment as a height flat interval. Through logical scanning of the entire time series, the indices of all time points conforming to the deviation characteristics and the indices of time points conforming to the flat characteristics are recorded, resulting in the phase deviation interval index and the height flat interval index.

[0024] The phase correction submodule is based on the phase deviation interval index and the height smooth interval index. It uses an interpolation fitting algorithm to reconstruct the corresponding deviation interval position in the carrier phase difference sequence. After point-by-point offset comparison between the repeated acquisition data and the original data, the data is replaced to obtain a continuous phase sequence. First, the index positions in the phase sequence where signal anomalies exist are located, such as index numbers 500 to 520. Based on the redundant observation mechanism established by the receiver front-end during the data acquisition phase, the multi-channel observation data or parallel observation results synchronously cached within this time window are called to obtain the carrier phase difference redundant observation sequence for the corresponding deviation interval, thereby forming highly reliable reference phase data. Subsequently, the phase values ​​in the redundant observation sequence are compared point by point with the damaged phase values ​​in the original sequence to calculate the phase offset deviation. Based on this deviation, a phase correction amount is constructed and fed back to the carrier tracking loop to perform centimeter-level phase alignment processing on the abnormal phase points to maintain carrier coherence. Further, when the phase offset deviation is logically determined to meet the non-physical cycle slip characteristics, the corresponding data point is determined to be an abnormal phase point, and the abnormal point is replaced and corrected using the redundant observation results or the estimated value of the adjacent time-stable interval. Finally, point-by-point continuity correction is performed on the corrected phase sequence, and phase expansion processing is performed in combination with the cycle slip repair number and the carrier wavelength to eliminate the discontinuity caused by phase jumps, resulting in a phase continuous sequence.

[0025] Specifically, such as Figure 2 , 4 As shown, the angle reconstruction module includes: The incident angle extraction submodule is based on the phase continuous sequence to obtain the incident angle and propagation time of multi-frequency signals at the end of the tower crane boom and the column structure of the lifting operation platform. It extracts the incident angle change and compares the incident angle data with the angle range of the structural boundary point by point. It removes the incident angle data that exceeds the structural boundary range to obtain the angle change amplitude sequence. First, the spatial spectrum estimation processing unit is invoked. A multi-signal classification algorithm is used to analyze the phase gradient distribution of the phase continuous sequence on the antenna array. By searching the spectral peak function, the incident angle of the signal arriving at the receiving antenna is calculated to be 55.6 degrees, and the signal propagation time is simultaneously extracted as 145.2 nanoseconds. This is then used in conjunction with the angular resolution parameters required for high-precision positioning for spatial mapping. Subsequently, the change in incident angle is extracted by comparing the angle values ​​of adjacent epochs. For example, subtracting the previous epoch's angle of 55.2 degrees from the current angle of 55.6 degrees yields a change of 0.4 degrees. To ensure the physical authenticity of the data, the extracted incident angle data is compared point-by-point with the pre-mapped structural boundary angle range. This boundary range is set to 10 to 170 degrees based on the tower crane's operating radius and the column's obstruction area. If a calculated incident angle is 175 degrees, it is determined to be an out-of-bounds signal and is discarded, retaining only valid observations within the physical structural boundaries. This allows for the filtering out of non-line-of-sight multipath interference through a geometric envelope within the high-precision positioning framework, resulting in the angle change amplitude sequence.

[0026] The direction association submodule maps the incident angle data at time nodes in the angle change amplitude sequence to the corresponding propagation time direction based on the time sequence correspondence, and determines the direction pointing change according to the time sequence correspondence to obtain the direction pointing sequence. First, the incident angle value corresponding to each sampling moment in the angle change amplitude sequence is extracted. For example, the incident angle at time point 12.5 is 62.8 degrees. Simultaneously, the propagation time direction vector corresponding to this time point is obtained. This direction is a spatial unit vector constructed using the electromagnetic wave's time of flight. This unit vector is aligned with the coordinate system reference for high-precision positioning, and a correspondence is established based on the chronological order of the timestamps. The trend of the incident angle change is then fused with the spatial propagation direction for calculation. For example, when the incident angle increases by 0.5 degrees, the corresponding propagation direction vector deflects clockwise within the horizontal reference plane. By traversing all time points, isolated angle changes are converted into dynamic vectors with clear spatial directions, thereby determining the continuous evolution of the direction direction in three-dimensional space and obtaining the direction direction sequence.

[0027] The consistency determination submodule is based on the direction pointing sequence. It compares the direction pointing of adjacent time nodes point by point. By comparing the range of direction angle change with the preset direction offset threshold, it identifies the direction change offset segment and corrects the direction to obtain the path direction sequence. First, adjacent time nodes in the direction pointing sequence are extracted, and their direction pointing deflection angles are compared point by point. For example, comparing the direction pointing at 145.2 degrees at time point 20 with the direction pointing at 146.5 degrees at time point 21, the calculated direction angle change range is 1.3 degrees. Then, this change range is compared with a preset direction offset threshold. This preset direction offset threshold is set at 25 degrees based on the physical limit of the turning radius of high-altitude walking gait, serving as a logical red line for the rationality of gait in high-precision positioning. If the instantaneous turning angle of a certain node reaches 45 degrees, exceeding the preset direction offset threshold of 20 degrees, this segment is identified as a direction change offset segment. For this offset segment, the preceding smoothing vector is called to perform direction correction, and the abnormal deflection points are forcibly calibrated onto the motion trajectory using a five-point cubic smoothing method, resulting in a path direction sequence.

[0028] Specifically, such as Figure 2 , 5 As shown, the trajectory generation module includes: The trajectory sorting submodule obtains the corresponding time node positioning coordinate sequence and height change sequence based on the path direction sequence, sorts the positioning coordinate sequence according to the timestamp order and extracts the trajectory point sequence, analyzes the temporal continuity of the trajectory point sequence, and obtains the trajectory point time sequence. First, the positioning coordinate sequence for each time node is acquired, including eastward and northward coordinates, as well as the altitude change sequence calculated jointly by barometer and phase analysis. The resolution of each coordinate component is then standardized to the millimeter-level range required for high-precision positioning. The positioning coordinate sequences distributed in the storage array are sorted in ascending order according to the atomic-level timestamps to ensure temporal linearity for each coordinate point. Subsequently, the trajectory point sequence is extracted from the sorted data stream, and its temporal continuity is analyzed. For example, the time difference between sequence numbers 200 and 201 is checked to see if it equals the preset sampling period of 0.05 seconds. If a time difference of 0.15 seconds is found, it is determined that there is a sampling gap or link interruption. By performing a temporal integrity check on all trajectory points, a spatial point cloud with rigorous temporal logic is established, resulting in the trajectory point time series sequence.

[0029] The connection determination submodule identifies the connection status based on the trajectory point time sequence, compares the spatial position and time interval of adjacent trajectory points, and obtains the trajectory interruption position sequence according to the spatial interval threshold and the time interval threshold. First, the spatial coordinates of adjacent trajectory points in the time series are extracted, and the Euclidean distance between them is calculated as the spatial interval. The time interval between the two points is also recorded. For example, the spatial interval between point A and point B is extracted to be 0.15 meters, and the time interval is 0.05 seconds. Then, the obtained values ​​are compared with preset spatial interval thresholds of 0.5 meters and time interval thresholds of 0.1 seconds. If the spatial interval exceeds the threshold, the current trajectory is determined to be disconnected and identified as a connection anomaly. The trajectory jump point is then located using high-precision positioning spatial domain association logic. By traversing and comparing the entire time series, the indexes of all logical breaks are recorded to determine which point pairs have trajectory jumps or losses, thus obtaining the trajectory interruption position sequence.

[0030] The trajectory matching submodule interpolates and supplements the trajectory interruption position sequence and compares it point by point with the height change sequence. The spatial changes of the supplemented trajectory points are consistent with the original trajectory points, thus obtaining a three-dimensional trajectory structure sequence. First, interpolation is performed to fill in the missing segments of the trajectory interruption sequence. The velocity and acceleration vectors before and after the interruption point are extracted, and a cubic spline interpolation algorithm is used to generate the spatial coordinates of the supplemented trajectory points. Then, these supplemented trajectory points are compared point-by-point with the synchronously acquired height change sequence to verify whether the horizontal displacement and height increase conform to the physical slope of the tower crane structure, ensuring that the completed trajectory still meets the geometric constraints of high-precision positioning in three-dimensional space. For example, the comparison revealed that the spatial change of the supplemented trajectory points is 0.85 meters, consistent with the historical movement slope of the original trajectory points, with a deviation rate of less than 5%, thus obtaining the three-dimensional trajectory structure sequence.

[0031] Specifically, such as Figure 2 , 6 As shown, the spindle weight module includes: The difference extraction submodule is based on the three-dimensional trajectory structure sequence. It obtains the coordinates of multiple sources at the same time node, compares the spatial position of the multi-source trajectory coordinates point by point, extracts the directional position difference based on the coordinate axis difference, and matches them according to the time node order to obtain the position difference sequence. First, the trajectory coordinates generated by different sensor sources at the same time point are obtained, for example, the BeiDou positioning coordinates (120.5, 45.8, 90.2) and the ultra-wideband assisted positioning coordinates (120.6, 45.7, 90.4). The execution spatial position of these multi-source trajectory coordinates is compared point by point, and the directional position difference is extracted by calculating the algebraic difference along each coordinate axis. This difference sequence reflects the convergence degree of high-precision positioning in a multipath environment. Specifically, the eastward difference is calculated to be 0.1 meters, the northward difference to be -0.1 meters, and the vertical difference to be 0.2 meters. Subsequently, these difference values ​​are precisely correlated according to the time node sequence to form a difference curve, thus obtaining the position difference sequence.

[0032] The direction recognition submodule extracts the magnitude of the direction position difference based on the position difference sequence, compares the magnitude data with the error magnitude threshold item by item, classifies the direction category according to the range of magnitude change, and obtains the direction category sequence corresponding to the time node. First, the overall amplitude of the directional position difference is extracted. By performing a square root operation on the sum of squares of the differences in each axis, the instantaneous position difference amplitude is found to be 0.24 meters. Next, this amplitude data is compared item by item with a preset error amplitude threshold sequence. This threshold sequence is set in three levels: within 0.15 meters is the excellent category, 0.15 to 0.45 meters is the acceptable category, and more than 0.45 meters is the out-of-tolerance category. This category classification directly maps to the quality level of high-precision positioning. The directional category is determined based on the range within which the amplitude change falls; for example, 0.24 meters is marked as category number 2. Subsequently, the category determination is correlated with time points to identify the reliability status of the trajectory at different stages, resulting in a directional category sequence.

[0033] The weight allocation submodule assigns weights to each direction category based on the direction category sequence, associates and maps the direction relationships, compares the continuity of the direction relationships between adjacent time nodes, identifies the offset direction based on the range of direction changes, and obtains the trajectory direction weight sequence. First, weights are assigned to each directional category according to a preset mapping table. The weight for the excellent category is 0.92, the weight for the qualified category is 0.65, and the weight for the out-of-tolerance category is 0.15. Then, the directional relationships between different time nodes are associated and mapped. By comparing the continuity of the directional relationships between adjacent time nodes, the offset direction is identified by analyzing the range of directional changes, and it is determined whether the direction is affected by environmental interference. For directions determined to be affected by interference, their weight ratio in the high-precision positioning fusion calculation is automatically reduced, and the remaining weights are redistributed to observation directions with higher weights to ensure that the overall weight sum remains at 1, finally obtaining the trajectory direction weight sequence.

[0034] Specifically, such as Figure 2 , 7 As shown, the trajectory analysis module includes: The state extraction submodule extracts the trajectory change state at time nodes based on the trajectory direction weight sequence, and maps the direction weights of time nodes in the trajectory direction weight sequence to the corresponding trajectory position changes to obtain the trajectory change sequence. First, the trajectory change state at each time node is extracted. This is achieved by coupling the time node direction weights in the trajectory direction weight sequence with the corresponding trajectory position change rate and height change rate. For example, at the 850th time node, the direction weight is 0.88, the horizontal position change rate is 1.2 m / s, and the height change rate is 0.5 m / s. These parameters are constructed into a multi-dimensional feature vector, which is then combined with preset motion state judgment rules to identify the movement state of personnel. When both the horizontal and vertical change rates are at a high level, it is determined to be a continuous movement state, while when the change rate is close to zero or below a preset threshold, it is determined to be a stationary or standby state, thus avoiding unrealistic state judgments. Furthermore, through joint analysis of the trajectory direction weights and position change features, different movement states are distinguished, resulting in a trajectory change sequence.

[0035] The trend comparison submodule is based on the trajectory change sequence. It compares the trajectory changes at adjacent time nodes with the range of directional changes and a preset change benchmark to identify discontinuous change nodes and obtain a set of change nodes. First, the motion characteristics of adjacent time nodes in the trajectory change sequence are extracted, and the changing trends of their velocity and direction vectors are compared point by point. A comparative analysis is performed based on a preset directional change range of 30 degrees and a preset acceleration change benchmark of 1.5 meters per second squared. For example, between time points 120 and 121, if the velocity suddenly increases to 8.5 meters per second and the direction deviates, its instantaneous acceleration is calculated and compared with the benchmark value. It is determined whether this change exceeds the limits of human biomechanics. By filtering such data abrupt changes, high-precision positioning output is ensured to be within the limits of physical laws, resulting in a set of change nodes.

[0036] The association matching submodule is based on the set of changing nodes, associates discontinuous time nodes, reorganizes the time order of nodes in the set of changing nodes, and compares it with the continuity of the overall trajectory change to obtain the trajectory behavior association structure; First, the time nodes marked as discontinuous are associated, and the trajectory is reconstructed through the consistency of behavioral logic. The time node sequence in the set of changing nodes is extracted, and logical verification is performed based on the historical movement habits of personnel and the connection relationship with the physical structure of the tower crane. For example, the climbing states before and after the break are compared on the time axis, and the continuity characteristics of the overall trajectory change are compared. Global topological constraints are used to compensate for the trajectory breakage in the signal blind zone of high-precision positioning. If the reconstructed behavioral curve conforms to the logical flow, the behavior repair and association of the nodes are completed, and the trajectory behavior association structure is obtained.

[0037] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A multi-scenario high-altitude worker positioning and trajectory data fusion analysis system, characterized in that, The system includes: The phase acquisition module acquires the carrier phase difference sequence and barometer height value sequence received by the terminal worn by the high-altitude worker, extracts the carrier phase difference change and height value change at adjacent moments, replaces the deviation value according to the offset, and obtains a continuous phase sequence. Based on the phase continuity sequence, the angle reconstruction module obtains the incident angle and propagation time at the end of the tower crane boom and the column structure of the lifting platform, compares it with the structural boundary angle range and associates it with the propagation direction to obtain the path direction sequence. Based on the path direction sequence, the trajectory generation module obtains the time node positioning coordinate sequence and the height change sequence, analyzes the connection status and supplements the interruption position matching height change sequence to obtain the three-dimensional trajectory structure sequence. Based on the three-dimensional trajectory structure sequence, the main axis weighting module extracts the multi-source trajectory coordinates and position differences at time nodes, compares them with the error amplitude threshold, assigns directional weights and associates them with directions to obtain the trajectory directional weight sequence. Based on the trajectory direction weight sequence, the trajectory analysis module extracts the trajectory change status at time nodes, compares the change trends of adjacent nodes, and associates discontinuous nodes to match the time sequence, thereby obtaining the trajectory behavior association structure.

2. The multi-scenario high-altitude worker positioning and trajectory data fusion analysis system according to claim 1, characterized in that, The phase continuity sequence includes a set of phase continuity values, jump correction markers, time index labels, height constraint labels, and offset correction parameters. The path direction sequence includes a set of direction vectors, incident angle identifiers, propagation time labels, direction consistency markers, and boundary angle constraints. The three-dimensional trajectory structure sequence includes a set of spatial coordinate points, a timestamp sequence, height level identifiers, trajectory connection relationships, and completion node markers. The trajectory direction weight sequence includes direction weight values, error amplitude labels, direction category labels, continuity scores, and weight distribution parameters. The trajectory behavior association structure includes behavior state labels, change trend indexes, discontinuous node markers, rearranged time chains, and association relationships.

3. The multi-scenario high-altitude worker positioning and trajectory data fusion analysis system according to claim 1, characterized in that, The aforementioned replacement deviation value refers to replacing the detected abnormal deviation data with re-acquired and corrected data by comparing the degree of data deviation. The reference structural boundary angle range refers to comparing the measured incident angle with the preset angle range of the preset structure and eliminating invalid angle data that exceeds the structural constraints.

4. The multi-scenario high-altitude worker positioning and trajectory data fusion analysis system according to claim 1, characterized in that, The trend of adjacent node changes refers to the continuous analysis of the direction and magnitude of changes of adjacent nodes in the time series; The discontinuous node refers to a trajectory node that has abrupt changes and disrupts continuity with the preceding and following nodes in terms of temporal and spatial variations.

5. The multi-scenario high-altitude worker positioning and trajectory data fusion analysis system according to claim 1, characterized in that, The phase acquisition module includes: The phase change extraction submodule acquires the carrier phase difference sequence and barometer height value sequence received by the terminal worn by the high-altitude worker, extracts the carrier phase difference change and height value change at adjacent moments, compares the differences between adjacent data in the carrier phase difference sequence, compares the differences between adjacent data in the barometer height value sequence, and obtains the phase change amplitude and height change amplitude. The interval identification submodule, based on the phase change amplitude and the height change amplitude, compares the phase change amplitude with a preset phase jump threshold to identify the deviation interval, and compares the height change amplitude with a preset height change threshold to identify the smooth interval, thus obtaining the phase deviation interval index and the height smooth interval index. The phase correction submodule, based on the phase deviation interval index and the height smooth interval index, uses an interpolation fitting algorithm to reconstruct the corresponding deviation interval position in the carrier phase difference sequence, and replaces the repeated acquisition data with the original data after point-by-point offset comparison, to obtain a continuous phase sequence.

6. The multi-scenario high-altitude worker positioning and trajectory data fusion analysis system according to claim 1, characterized in that, The angle reconstruction module includes: The incident angle extraction submodule, based on the phase continuity sequence, obtains the incident angle and propagation time of the multi-frequency signal at the end of the tower crane boom and the column structure of the lifting operation platform, extracts the incident angle change, and compares the incident angle data with the angle range of the structural boundary point by point, removes the incident angle data that exceeds the structural boundary range, and obtains the angle change amplitude sequence. The direction association submodule maps the incident angle data at time nodes in the angle change amplitude sequence to the corresponding propagation time direction based on the angle change amplitude sequence, and determines the direction pointing change according to the time sequence correspondence to obtain the direction pointing sequence. The consistency determination submodule, based on the direction pointing sequence, compares the direction pointing of adjacent time nodes point by point, identifies the direction change offset segment by comparing the range of direction angle change with the preset direction offset threshold, and corrects the direction to obtain the path direction sequence.

7. The multi-scenario high-altitude worker positioning and trajectory data fusion analysis system according to claim 1, characterized in that, The trajectory generation module includes: Based on the path direction sequence, the trajectory sorting submodule obtains the corresponding time node positioning coordinate sequence and height change sequence, sorts the positioning coordinate sequence according to the timestamp order and extracts the trajectory point sequence, analyzes the temporal continuity of the trajectory point sequence, and obtains the trajectory point time sequence. The connection determination submodule identifies the connection status based on the trajectory point time sequence, compares the spatial position and time interval of adjacent trajectory points, and obtains the trajectory interruption position sequence according to the spatial interval threshold and the time interval threshold. The trajectory matching submodule interpolates and supplements the trajectory interruption position sequence and compares it point by point with the height change sequence. The spatial changes of the supplemented trajectory points are consistent with the original trajectory points, thus obtaining a three-dimensional trajectory structure sequence.

8. The multi-scenario high-altitude worker positioning and trajectory data fusion analysis system according to claim 1, characterized in that, The spindle weighting module includes: The difference extraction submodule obtains the coordinates of multiple sources at the same time node based on the three-dimensional trajectory structure sequence, compares the spatial position of the multi-source trajectory coordinates point by point, extracts the directional position difference according to the coordinate axis difference, and matches them according to the time node order to obtain the position difference sequence. The direction recognition submodule extracts the magnitude of the direction position difference based on the position difference sequence, compares the magnitude data with the error magnitude threshold item by item, classifies the direction category according to the range of magnitude change, and obtains the direction category sequence corresponding to the time node. The weight allocation submodule assigns weights to the direction category based on the direction category sequence, associates and maps the direction relationships, compares the continuity of the direction relationships between adjacent time nodes, identifies the offset direction based on the range of direction changes, and obtains the trajectory direction weight sequence.

9. The multi-scenario high-altitude worker positioning and trajectory data fusion analysis system according to claim 1, characterized in that, The trajectory analysis module includes: The state extraction submodule extracts the trajectory change state at time nodes based on the trajectory direction weight sequence, and maps the direction weights of time nodes in the trajectory direction weight sequence to the corresponding trajectory position changes to obtain the trajectory change sequence. The trend comparison submodule, based on the trajectory change sequence, compares the trajectory changes at adjacent time nodes, and compares the range of directional changes with a preset change benchmark to identify discontinuous change nodes and obtain a set of change nodes. The association matching submodule, based on the set of changing nodes, associates discontinuous time nodes, reorganizes the time order of nodes in the set of changing nodes, and compares the overall trajectory change continuity to obtain the trajectory behavior association structure.

10. The multi-scenario high-altitude worker positioning and trajectory data fusion analysis system according to claim 9, characterized in that, In the process of matching the time node direction weight with the corresponding trajectory position change: the time interval between adjacent time nodes is limited to a preset time window for matching one by one, and a sliding comparison is performed on three consecutive time nodes in the trajectory direction weight sequence; In the process of reorganizing the time order of nodes in the set of changed nodes: the nodes are sorted from earliest to latest according to the timestamp, and the time interval between adjacent time nodes after sorting is checked for continuity. Time nodes whose time interval exceeds the preset time window are removed.

Images