A Method and System for Campus Main Trajectory Analysis Based on Multi-Source Fusion Data

Abstract

This invention relates to the field of smart campuses, providing a method and system for analyzing the main trajectory of campus traffic using multi-source fusion data. The method includes the following steps: synchronously collecting raw vehicle data via a smart cycling terminal, including triaxial acceleration data, tire pressure data, and high-precision positioning trajectory data; preprocessing the raw data and extracting acceleration mutation events, tire pressure drop events, and trajectory deceleration features; spatially overlaying the preprocessed event data with a campus road network GIS layer, generating a pothole damage heatmap based on preset damage assessment rules; extracting high-risk road sections based on the pothole damage heatmap and pushing high-risk road section warnings to users. This invention provides a scientific basis for campus road maintenance planning, enabling preventative maintenance and reducing road surface damage.

Description

Technical Field

[0001] This invention relates to the field of smart campuses, specifically to a method and system for analyzing the trajectory of main campus entities using multi-source fusion data. Background Technology

[0002] Due to factors such as geological subsidence, improper backfilling during construction, and long-term vehicle traffic, roads within the campus are prone to surface depressions, cracks, and other potholes. These types of damage are typically small in area, highly concealed, but extremely dangerous, easily leading to tire blowouts, skidding, and other safety accidents involving bicycles, electric bikes, and other vehicles. Traditional maintenance methods rely on regular manual inspections or complaints from teachers and students, but these methods are inefficient, resulting in many small potholes remaining undetected for extended periods.

[0003] To improve detection efficiency, existing technologies utilize the deployment of fixed road monitoring equipment (such as cameras) to monitor key road sections, but this requires the pre-installation or erection of dedicated hardware facilities on campus roads.

[0004] However, the aforementioned solutions have fundamental flaws in campus scenarios. Fixed monitoring equipment is expensive and has limited coverage, making it difficult to achieve dynamic monitoring of the entire campus road network. Furthermore, some uneven road surfaces may be difficult to identify, requiring further investment in monitoring equipment to achieve accurate identification. Therefore, a method and system for analyzing the main trajectory of campus subjects using multi-source fusion data is proposed to address these issues. Summary of the Invention

[0005] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a method and system for analyzing the main trajectory of campus subjects using multi-source fusion data, so as to solve the problems existing in the above-mentioned background technology.

[0006] This invention is implemented as follows: a method for analyzing the main trajectory of a campus using multi-source fusion data, the method comprising the following steps:

[0007] The vehicle's raw data is collected synchronously through an intelligent cycling terminal, including triaxial acceleration data, tire pressure data, and high-precision positioning trajectory data. The intelligent cycling terminal is suitable for authorized non-motorized vehicles within the campus.

[0008] The raw data was preprocessed and acceleration mutation events, tire pressure drop events, and trajectory deceleration features were extracted.

[0009] The preprocessed event data is spatially overlaid with the campus road network GIS layer, and a pothole damage heat map is generated based on the preset damage judgment rules.

[0010] High-risk road sections are identified based on the pothole damage heat map, and warnings of high-risk road sections are pushed to users, while work orders are sent to the maintenance department.

[0011] As a further aspect of the present invention: the intelligent cycling terminal integrates a multimodal sensor group, which includes a triaxial accelerometer for detecting vertical impact acceleration, a tire pressure sensor for monitoring tire pressure changes, a trajectory recording unit supporting GPS positioning, a solar power supply unit for power supply, and a low-power communication unit. The low-power communication unit only uploads data when the acceleration peak or the tire pressure change rate is greater than a set value.

[0012] As a further aspect of the present invention: the step of preprocessing the raw data and extracting acceleration abrupt events, tire pressure drop events, and trajectory deceleration features specifically includes:

[0013] The triaxial acceleration data is processed by sliding window filtering to identify instantaneous impact characteristics in the vertical direction. When the impact intensity exceeds the preset threshold for normal road surface bumps, it is marked as an acceleration mutation event.

[0014] Based on real-time monitoring of tire pressure data, when a non-linear, precipitous drop in pressure is detected, it is confirmed as a sudden drop in tire pressure event.

[0015] The velocity change rate of continuous trajectory points is calculated based on high-precision positioning trajectory data. When the vehicle speed suddenly decreases significantly and continues for a specific period, it is extracted as a trajectory deceleration feature.

[0016] Establish a spatiotemporal correlation index for three types of events, and add timestamps and spatial confidence labels to the event data to filter out coordinate anomalies caused by satellite positioning drift.

[0017] As a further aspect of the present invention: the step of spatially overlaying the preprocessed event data with the campus road network GIS layer and generating a pothole damage heat map based on preset damage determination rules specifically includes:

[0018] The spatial coordinates of acceleration mutation events, tire pressure drop events, and trajectory deceleration characteristics are mapped to the campus road network GIS layer to construct a spatial aggregation model of event density distribution.

[0019] Based on the campus road topology, a dynamic adaptive grid is divided, and the cumulative event frequency within the grid is weighted and statistically analyzed, with the sudden tire pressure drop event given the highest weight coefficient.

[0020] Access the whitelist database of inherent infrastructure locations on the road to filter acceleration abrupt change interference events caused by inherent infrastructure.

[0021] A damage probability heatmap is generated based on the event weight statistics, and a gradient color scale is used to represent the severity of the damage, with the high-frequency tire pressure event area rendered as the highest risk level.

[0022] By integrating historical maintenance records, the thermal values ​​of repaired areas are attenuated and suppressed, thereby enabling dynamic evolution of the reliability of the heat map.

[0023] As a further aspect of the present invention: the step of calling the whitelist database of inherent facility locations on the road to filter acceleration abrupt interference events caused by inherent facilities specifically includes:

[0024] A multi-vehicle speed bump vibration feature library is constructed, and the acceleration waveform features when passing over the speed bump at different speeds are extracted by machine learning. The acceleration waveform features include the amplitude envelope shape, spectral energy distribution and duration.

[0025] When an acceleration mutation event is detected, the dynamic time warping distance between its waveform and all templates in the feature library is calculated in real time. If the minimum distance is lower than the adaptive matching threshold, it is determined to be an inherent facility interference event.

[0026] By combining the GIS coordinate data of the existing facilities with the vehicle's direction of travel, a three-dimensional spatial influence cone model is established, so that the filtering mechanism is triggered only when the acceleration mutation event simultaneously satisfies waveform matching and is located within the influence cone range;

[0027] We continuously collect data on newly occurring acceleration mutation events. When similar events that do not match the feature library form significant clusters after spatial clustering, we automatically generate learning samples that are suspected to be new speed bumps and request manual annotation.

[0028] As a further aspect of the present invention: the step of extracting high-risk road sections based on pothole damage heatmaps and pushing high-risk road section warnings to users specifically includes:

[0029] The user's behavioral data is collected through a smart cycling terminal, including cycling speed and route information;

[0030] Spatial correlation matching is performed between the distribution characteristics of high-risk road sections based on pothole damage heatmaps and user behavior data;

[0031] When a user approaches a high-risk section of the heat map, a prompt instruction is sent to the smart cycling terminal, causing the smart cycling terminal to emit a warning sound and light. The smart cycling terminal also integrates an audible and visual alarm.

[0032] Another objective of this invention is to provide a campus main trajectory analysis system based on multi-source fusion data, the system comprising:

[0033] The data synchronization acquisition module is used to synchronously collect the vehicle's raw data through the intelligent cycling terminal. The raw data includes triaxial acceleration data, tire pressure data, and high-precision positioning trajectory data. The intelligent cycling terminal is suitable for authorized non-motorized vehicles within the campus.

[0034] The event extraction module is used to preprocess the raw data and extract acceleration mutation events, tire pressure drop events, and trajectory deceleration features;

[0035] The heat map generation module is used to spatially overlay the preprocessed event data with the campus road network GIS layer and generate a pothole damage heat map based on preset damage judgment rules.

[0036] The linkage response module is used to extract high-risk road sections based on the pothole damage heat map, push high-risk road section warnings to users, and simultaneously output work orders to the maintenance department.

[0037] As a further aspect of the present invention: the event extraction module includes:

[0038] The acceleration filtering unit is used to perform sliding window filtering on triaxial acceleration data, identify instantaneous impact characteristics in the vertical direction, and mark an acceleration mutation event when the impact intensity exceeds the preset conventional road bump threshold.

[0039] The tire pressure monitoring unit is used to monitor the tire pressure change trend in real time based on tire pressure data. When a non-linear, cliff-like drop in pressure value is detected, it is confirmed as a sudden drop in tire pressure event.

[0040] The speed decay extraction unit is used to calculate the speed change rate of continuous trajectory points based on high-precision positioning trajectory data. When the vehicle speed suddenly decreases significantly and continues for a specific period, it is extracted as a trajectory deceleration feature.

[0041] The event association unit is used to establish spatiotemporal association indexes for three types of events, add timestamps and spatial confidence labels to event data, and filter out coordinate anomalies caused by satellite positioning drift.

[0042] As a further aspect of the present invention: the heat map generation module includes:

[0043] The event space mapping unit is used to map the spatial coordinates of acceleration mutation events, tire pressure drop events, and trajectory deceleration characteristics to the campus road network GIS layer, and to construct a spatial aggregation model of event density distribution.

[0044] The weighted statistical unit is used to divide the dynamic adaptive grid according to the campus road topology and perform weighted statistics on the cumulative event frequency within the grid, with the sudden tire pressure drop event given the highest weight coefficient.

[0045] The interference filtering unit is used to call the whitelist database of the location of inherent facilities on the road to filter interference events caused by acceleration abrupt changes due to inherent facilities;

[0046] The thermal rendering output unit is used to generate a damage probability heatmap based on the event weight statistics results. Gradient color levels are used to represent the severity of the damage, with the high-frequency tire pressure event area rendered as the highest risk level.

[0047] The dynamic attenuation suppression unit is used to integrate historical maintenance record data to attenuate and suppress the thermal values ​​of the repaired area, thereby enabling dynamic reliability evolution of the heat map.

[0048] As a further aspect of the present invention: the linkage response module includes:

[0049] A behavior data acquisition unit is used to collect user behavior data through a smart cycling terminal, the behavior data including cycling speed and route information;

[0050] The thermal path matching unit is used to perform spatial correlation matching between the distribution characteristics of high-risk road sections based on the pothole damage heat map and user behavior data.

[0051] The warning triggering unit is used to send a prompt command to the smart cycling terminal when the user approaches a high-risk section of the heat map, causing the smart cycling terminal to emit a warning sound and light. The smart cycling terminal also integrates an sound and light alarm.

[0052] Compared with the prior art, the beneficial effects of the present invention are:

[0053] This invention utilizes intelligent cycling terminals deployed on campus-owned non-motorized vehicles to simultaneously collect multi-source data, including high-precision trajectory, three-axis acceleration, and tire pressure. It integrates and analyzes the spatiotemporal correlation between acceleration abrupt changes, sudden tire pressure drops, and trajectory deceleration characteristics to construct a dynamic damage heatmap. This effectively addresses the technical pain points of low efficiency in traditional manual inspections and the high cost and limited coverage of fixed monitoring equipment. Compared to existing image monitoring solutions with low recognition rates for hidden potholes, this solution innovatively introduces sudden tire pressure drops as a physical damage standard. By monitoring abnormal changes in tire pressure in real time, it can accurately detect road surface depressions, achieving higher-precision pothole detection across the entire campus road network without large-scale infrastructure investment. It can intelligently push obstacle avoidance warnings to teachers and students and provide work orders containing precise location coordinates and damage levels to maintenance departments, forming a complete closed loop from damage identification and early warning to repair acceptance. It can automatically execute monitoring tasks around the clock, significantly improving the timeliness and economy of campus road safety management. Furthermore, through long-term data accumulation and analysis, we can provide a scientific basis for campus road maintenance planning, achieve preventive maintenance, and reduce road surface damage from the source. Attached Figure Description

[0054] Figure 1 This is a flowchart of a method for analyzing the main trajectory of campus data using multi-source fusion.

[0055] Figure 2 This is a flowchart illustrating the preprocessing of raw data and extraction of acceleration mutation events, tire pressure drop events, and trajectory deceleration characteristics in a campus main trajectory analysis method based on multi-source fusion data.

[0056] Figure 3 The flowchart describes a method for analyzing the main trajectory of campus data based on multi-source fusion. It involves spatially overlaying preprocessed event data with a campus road network GIS layer and generating a pothole damage heat map based on preset damage judgment rules.

[0057] Figure 4 This is a flowchart illustrating how a whitelist database of inherent facility locations on roads is used to filter acceleration abrupt interference events caused by these inherent facilities in a campus main trajectory analysis method based on multi-source fusion data.

[0058] Figure 5 This is a flowchart illustrating the process of extracting high-risk road sections based on pothole damage heatmaps and pushing high-risk road section warnings to users in a campus main trajectory analysis method using multi-source fusion data.

[0059] Figure 6 This is a schematic diagram of the structure of a campus main trajectory analysis system based on multi-source fusion data.

[0060] Figure 7 This is a schematic diagram of the event extraction module in a campus main trajectory analysis system based on multi-source fusion data.

[0061] Figure 8 This is a schematic diagram of the heatmap generation module in a campus main trajectory analysis system based on multi-source fusion data.

[0062] Figure 9 This is a schematic diagram of the linkage response module in a campus main trajectory analysis system based on multi-source fusion data. Detailed Implementation

[0063] 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 specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0064] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0065] like Figure 1 As shown in the figure, this invention provides a method for analyzing the trajectory of main campus data using multi-source fusion data. The method includes the following steps:

[0066] S100 synchronously collects raw data of the vehicle through an intelligent cycling terminal. The raw data includes triaxial acceleration data, tire pressure data, and high-precision positioning trajectory data. The intelligent cycling terminal is suitable for authorized non-motorized vehicles within the campus.

[0067] S200 preprocesses the raw data and extracts acceleration mutation events, tire pressure drop events, and trajectory deceleration features;

[0068] S300 spatially overlays the pre-processed event data with the campus road network GIS layer and generates a pothole damage heat map based on preset damage judgment rules.

[0069] The S400 identifies high-risk road sections based on a pothole damage heat map, sends warnings to users about these sections, and simultaneously generates work orders for the maintenance department.

[0070] It should be noted that the intelligent cycling terminal integrates a multimodal sensor group, which includes a triaxial accelerometer for detecting vertical impact acceleration, a tire pressure sensor for monitoring tire pressure changes, a trajectory recording unit supporting GPS positioning, a solar power supply unit for power supply, and a low-power communication unit. The low-power communication unit only uploads data when the acceleration peak or the tire pressure change rate is greater than a set value. For example, when a certain road section simultaneously triggers an acceleration surge peak exceeding 2.5 g or a tire pressure change rate higher than 20 kPa / s, data will be uploaded. In addition, the authorized non-motorized vehicles are generally electric bikes or bicycles on campus. Why not motorized vehicles? Because non-motorized vehicles can travel on various roads on campus, while motorized vehicles are more restricted. Using non-motorized vehicles allows for the collection of a wider range of data, facilitating comprehensive analysis.

[0071] In this embodiment of the invention, an intelligent cycling terminal deployed on campus-owned non-motorized vehicles simultaneously collects multi-source data such as high-precision trajectory, three-axis acceleration, and tire pressure. It integrates and analyzes the spatiotemporal correlation between acceleration abrupt changes, sudden tire pressure drops, and trajectory deceleration characteristics to construct a dynamic damage heatmap, effectively solving the technical pain points of low efficiency in traditional manual inspections and high cost and limited coverage of fixed monitoring equipment. Compared to existing image monitoring solutions with low recognition rates for hidden potholes, this solution innovatively introduces sudden tire pressure drops as a physical damage standard. By monitoring abnormal changes in tire pressure in real time, it can accurately detect road surface depressions, achieving higher-precision pothole detection across the entire campus road network without large-scale infrastructure investment. It can intelligently push obstacle avoidance warnings to teachers and students and provide work orders containing precise location coordinates and damage levels to the maintenance department, forming a complete closed loop from damage identification and early warning to maintenance acceptance. It can automatically execute monitoring tasks around the clock, significantly improving the timeliness and economy of campus road safety management. Furthermore, through long-term data accumulation and analysis, we can provide a scientific basis for campus road maintenance planning, achieve preventive maintenance, and reduce road surface damage from the source.

[0072] like Figure 2 As shown, in a preferred embodiment of the present invention, the step of preprocessing the raw data and extracting acceleration abrupt events, tire pressure drop events, and trajectory deceleration features specifically includes:

[0073] S201 performs sliding window filtering on the triaxial acceleration data to identify instantaneous impact characteristics in the vertical direction. When the impact intensity exceeds the preset threshold for normal road surface bumps, it is marked as an acceleration mutation event.

[0074] S202, based on real-time monitoring of tire pressure change trends using tire pressure data, confirms a sudden drop in tire pressure when a non-linear, cliff-like drop in pressure value is detected.

[0075] S203, calculates the rate of change of speed of continuous trajectory points based on high-precision positioning trajectory data, and extracts the trajectory deceleration feature when the vehicle speed suddenly decreases significantly and continues for a specific period.

[0076] S204. Establish a spatiotemporal correlation index for three types of events, add timestamps and spatial confidence labels to the event data, and filter out coordinate anomalies caused by satellite positioning drift.

[0077] In this embodiment of the invention, the steps of preprocessing the raw data and extracting feature events employ a multi-dimensional cross-validation technique, which demonstrates significant advantages in practical applications. When performing sliding window filtering on the triaxial acceleration data (e.g., using a 0.5-second time window), the system focuses on monitoring instantaneous impact characteristics in the vertical direction. This processing method can effectively distinguish between normal driving vibrations and abnormal pothole impacts. For example, when a vehicle travels at 15 km / h through a pothole with a diameter of 20 cm and a depth of 5 cm, the vertical acceleration detected by the system will be greater than 2.5 g, while ordinary road bumps are usually less than 1.5 g. In terms of tire pressure monitoring, special attention needs to be paid to nonlinear cliff-like drop characteristics (e.g., a tire pressure drop of more than 30 kPa within 3 seconds). This change pattern contrasts sharply with slow leaks (a drop of 2-3 kPa per hour) and can accurately identify tire blowout-level damage. In trajectory analysis, the rate of change of velocity at continuous trajectory points (e.g., 5 sampling points per second) is calculated. When a velocity drop of more than 40% within 2 seconds is detected and continues for more than 5 seconds, it can be determined as deceleration caused by avoiding or colliding with potholes. By establishing spatiotemporal correlation indexes for three types of events (e.g., data correlation within a 10-second time window and a 5-meter spatial range), and attaching timestamps and confidence labels, data anomalies caused by satellite positioning drift can be effectively filtered out, ensuring the accuracy of event detection. This multi-source data fusion processing method significantly reduces the false alarm rate compared to single-sensor solutions (e.g., using only accelerometers).

[0078] like Figure 3 As shown in the preferred embodiment of the present invention, the step of spatially overlaying the preprocessed event data with the campus road network GIS layer and generating a pothole damage heat map based on preset damage determination rules specifically includes:

[0079] S301 maps the spatial coordinates of acceleration mutation events, tire pressure drop events, and trajectory deceleration characteristics to the campus road network GIS layer to construct a spatial aggregation model of event density distribution.

[0080] S302, based on the campus road topology, divides the dynamic adaptive grid and performs weighted statistics on the cumulative event frequency within the grid, with the sudden tire pressure drop event given the highest weight coefficient.

[0081] S303, call the whitelist database of the location of inherent facilities on the road to filter acceleration change interference events caused by inherent facilities;

[0082] S304. A damage probability heatmap is generated based on the event weight statistics. Gradient color levels are used to represent the severity of the damage, with the high-frequency tire pressure event area rendered as the highest risk level.

[0083] S305 integrates historical maintenance record data to suppress the attenuation of thermal values ​​in repaired areas, thereby enabling dynamic reliability evolution of thermal maps.

[0084] In this embodiment of the invention, the process of generating a pothole damage heat map employs multi-level intelligent analysis technology. By accurately overlaying preprocessed event data with the campus GIS road network, precise spatial positioning and risk assessment of road surface damage are achieved. In specific implementation, the system first maps the spatial coordinates (latitude and longitude) of various events onto the campus GIS base map, constructing a spatial aggregation model based on kernel density estimation. This processing method can effectively identify event cluster areas. The system automatically adjusts the analysis grid size based on the actual road topology (assuming the main roads on campus are 6 meters wide and the side roads are 4 meters wide, then the main roads use a 5 m × 5 m grid and the side roads use a 3 m × 3 m grid). It performs differentiated weighted statistics on events within the grid, with tire pressure events, acceleration mutation events, and deceleration features having decreasing weights. By accessing a whitelist database of fixed facilities on the road (containing the precise coordinates of speed bumps, manhole covers, etc.), the system can intelligently filter out most interfering events (such as acceleration mutations within 2 meters of speed bumps). During the heatmap rendering stage, the system uses the HSV color space for gradient coloring; for example, blue, yellow, and red represent different areas, with high-frequency tire pressure event areas forcibly marked as red warning zones. It should be noted that it also integrates historical maintenance data from the past three months, applying exponential decay processing to the heatmap values ​​of these areas to ensure the heatmap dynamically reflects the latest road conditions. This processing method, which integrates spatial statistics, weighted analysis, and historical data, not only improves detection efficiency but also ensures recognition accuracy compared to traditional manual inspection methods.

[0085] like Figure 4 As shown, in a preferred embodiment of the present invention, the step of calling the whitelist database of inherent facility locations on the road to filter acceleration abrupt interference events caused by inherent facilities specifically includes:

[0086] S313, Construct a multi-vehicle speed bump vibration feature library, and extract acceleration waveform features when passing through speed bumps at different speeds through machine learning. The acceleration waveform features include amplitude envelope shape, spectral energy distribution and duration.

[0087] S323, when an acceleration mutation event is detected, the dynamic time warping distance between its waveform and all templates in the feature library is calculated in real time. If the minimum distance is lower than the adaptive matching threshold, it is determined to be an inherent facility interference event.

[0088] S333 combines the GIS coordinate data of the existing facilities with the vehicle's direction of travel to establish a three-dimensional spatial influence cone model, so that the filtering mechanism is triggered only when the acceleration mutation event simultaneously satisfies waveform matching and is within the influence cone range;

[0089] S343 continuously collects data on newly occurring acceleration mutation events. When similar events that do not match the feature library form significant clusters after spatial clustering, it automatically generates learning samples that are suspected to be newly added speed bumps and requests manual annotation.

[0090] In this embodiment of the invention, the intelligent filtering mechanism for inherent road infrastructure employs innovative multi-dimensional verification technology. By constructing a comprehensive speed bump feature library and intelligent matching algorithm, the accuracy of interference event filtering is significantly improved. Specifically, the system pre-establishes a speed bump vibration feature library containing bicycles and electric vehicles, and analyzes waveform characteristics in typical speed ranges (e.g., 15-20 km / h for electric vehicles, 10-15 km / h for bicycles) using machine learning. These characteristics include not only the amplitude envelope shape in the time domain but also the energy distribution characteristics and duration characteristics in the frequency domain. In actual operation, when an acceleration abrupt change event is detected, the system calculates the dynamic time warping (DTW) distance between its waveform and all templates in the feature library in real time. This algorithm effectively overcomes the waveform scaling problem caused by speed differences. For example, when the DTW distance of a certain acceleration abrupt change event has a similarity to the best matching template exceeding a preset matching threshold (e.g., 85%), it is determined to be speed bump interference. In addition, a three-dimensional spatial influence cone model is established by combining the precise GIS coordinates of the speed bump and the vehicle's direction of travel (calculated through trajectory angles). Events are only filtered if they simultaneously satisfy waveform matching and spatial location matching. It should be noted that when a sudden acceleration event with high spatial clustering that does not match the feature library is detected (such as 10 consecutive similar waveforms appearing within a 30-meter range on a certain road section), a learning sample suspected of being a new speed bump is automatically generated and pushed to the management backend to request manual annotation (e.g., confirmation through surveillance footage or on-site investigation), thus achieving dynamic updates to the feature library.

[0091] like Figure 5 As shown in the preferred embodiment of the present invention, the step of extracting high-risk road sections based on the pothole damage heat map and pushing high-risk road section warnings to users specifically includes:

[0092] S401, collects user behavior data through a smart cycling terminal, the behavior data including cycling speed and route information;

[0093] S402, Spatial correlation matching is performed between the distribution characteristics of high-risk road sections based on pothole damage heatmap and user behavior data;

[0094] S403 When a user approaches a high-risk section of the heat map, a prompt instruction is sent to the intelligent cycling terminal, causing the intelligent cycling terminal to emit a warning sound and light. The intelligent cycling terminal also integrates an sound and light alarm.

[0095] In this embodiment of the invention, the user's cycling behavior data, including speed change data and high-precision route trajectory, is first collected in real time through a smart cycling terminal. This data is intelligently matched and analyzed with a pothole damage heat map generated by the system. The heat map contains risk area markings of different levels. When the distance between the user's current location and a high-risk section is detected to be within the warning range (different areas have different warning ranges, with the high-risk section having the largest warning range), the warning timing is intelligently calculated based on the user's current speed, typically triggering a warning 10-15 seconds before reaching the danger point. The warning command is transmitted to the cycling terminal in real time through a low-power communication unit, and the terminal's built-in multi-mode alarm activates differentiated warnings based on the danger level: for ordinary risks (yellow areas), an intermittent buzzer (once per second, 60 decibels) is emitted, while for high-risk areas (red), a strong flashing light (100 lumens) is activated in conjunction with a continuous alarm (75 decibels). This ensures the timeliness and effectiveness of the warnings, reducing the accident rate for users on dangerous road sections. At the same time, it will also record users' avoidance behavior data, and this feedback data will be used to optimize the accuracy of the heat map, forming a virtuous cycle of continuous improvement.

[0096] like Figure 6 As shown in the figure, this embodiment of the invention also provides a campus main trajectory analysis system based on multi-source fusion data, the system comprising:

[0097] The data synchronization acquisition module 100 is used to synchronously collect the vehicle's raw data through the intelligent cycling terminal. The raw data includes triaxial acceleration data, tire pressure data, and high-precision positioning trajectory data. The intelligent cycling terminal is suitable for authorized non-motorized vehicles within the campus.

[0098] The event extraction module 200 is used to preprocess the raw data and extract acceleration mutation events, tire pressure drop events, and trajectory deceleration features;

[0099] The heat map generation module 300 is used to spatially overlay the preprocessed event data with the campus road network GIS layer and generate a pothole damage heat map based on preset damage judgment rules.

[0100] The linkage response module 400 is used to extract high-risk road sections based on the pothole damage heat map, push high-risk road section warnings to users, and output work orders to the maintenance department.

[0101] In this embodiment of the invention, the coordinated operation of four intelligent modules enables accurate identification and rapid response to road surface pothole damage. Specifically, the data synchronization acquisition module 100 collects triaxial acceleration, tire pressure, and high-precision positioning data in real time through an intelligent terminal (such as a dedicated device, model XC-2023) installed on electric vehicles or bicycles on campus. In practical applications, when a vehicle passes over a pothole at a speed of 15 km / h, this module can accurately record the vertical impact acceleration and instantaneous tire pressure drop. The event extraction module 200 performs noise reduction processing on the raw data, identifies characteristic events by setting multi-level thresholds (such as acceleration > 2.5 g, tire pressure drop > 20 kPa within 10 seconds, speed reduction > 40%), and establishes a spatiotemporal correlation index. The heat map generation module 300 overlays the processed data with a high-precision GIS map of the campus and generates a damage heat map using a kernel density estimation algorithm. For example, if five sudden tire pressure drops are detected within three days on a section of road in front of a teaching building, the system will mark that area as a red high-risk zone. The linkage response module 400 pushes early warning information in real time via the network. For example, when a student rides a bicycle and approaches within 50 meters of a dangerous area, the terminal will issue an audible and visual alarm and automatically generate a repair work order containing precise coordinates and damage photos. The efficiency of the entire system is significantly improved compared to the traditional manual inspection method.

[0102] like Figure 7 As shown, in a preferred embodiment of the present invention, the event extraction module 200 includes:

[0103] The acceleration filtering unit 201 is used to perform sliding window filtering on triaxial acceleration data, identify instantaneous impact characteristics in the vertical direction, and mark an acceleration mutation event when the impact intensity exceeds the preset conventional road bump threshold.

[0104] The tire pressure monitoring unit 202 is used to monitor the tire pressure change trend in real time based on tire pressure data. When a non-linear cliff drop in pressure value is detected, it is confirmed as a sudden drop in tire pressure event.

[0105] The speed decay extraction unit 203 is used to calculate the speed change rate of continuous trajectory points based on high-precision positioning trajectory data. When the vehicle speed suddenly decreases significantly and continues for a specific period, it is extracted as a trajectory deceleration feature.

[0106] Event association unit 204 is used to establish spatiotemporal association indexes for three types of events, add timestamps and spatial confidence labels to event data, and filter out coordinate anomalies caused by satellite positioning drift.

[0107] like Figure 8 As shown, in a preferred embodiment of the present invention, the heatmap generation module 300 includes:

[0108] Event space mapping unit 301 is used to map the spatial coordinates of acceleration mutation events, tire pressure drop events and trajectory deceleration characteristics to the campus road network GIS layer to construct a spatial aggregation model of event density distribution.

[0109] Weighted statistical unit 302 is used to divide a dynamic adaptive grid according to the campus road topology and perform weighted statistics on the cumulative event frequency within the grid, with the tire pressure drop event given the highest weight coefficient.

[0110] Interference filtering unit 303 is used to call the whitelist database of the location of inherent facilities on the road to filter acceleration change interference events caused by inherent facilities;

[0111] The thermal rendering output unit 304 is used to generate a damage probability heatmap based on the event weight statistics results. Gradient color levels are used to represent the severity of damage, with the high-frequency tire pressure event area rendered as the highest risk level.

[0112] The dynamic attenuation suppression unit 305 is used to integrate historical maintenance record data and suppress the attenuation of the thermal value of the repaired area to achieve dynamic reliability evolution of the heat map.

[0113] like Figure 9 As shown, in a preferred embodiment of the present invention, the linkage response module 400 includes:

[0114] The behavior data acquisition unit 401 is used to collect user behavior data through a smart cycling terminal, the behavior data including cycling speed and route information;

[0115] The thermal path matching unit 402 is used to perform spatial correlation matching between the distribution characteristics of high-risk road sections based on the pothole damage thermal map and user behavior data.

[0116] The warning triggering unit 403 is used to send a warning instruction to the smart cycling terminal when the user approaches a high-risk section of the heat map, causing the smart cycling terminal to emit a warning sound and light. The smart cycling terminal also integrates a sound and light alarm.

[0117] The above description only details the preferred embodiments of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0118] It should be understood that although the steps in the flowcharts of the various 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 the various embodiments 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.

[0119] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the disclosure in the specification and embodiments. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and embodiments are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the claims.

Claims

1. A method for campus subject trajectory analysis of multi-source fused data, characterized in that, The method comprises the following steps: Synchronize the original data of the vehicle through the intelligent riding terminal, the original data including three-axis acceleration data, tire pressure data and high-precision positioning trajectory data, and the intelligent riding terminal being suitable for authorized non-motor vehicles inside the campus; Preprocess the original data and extract acceleration mutation events, tire pressure sudden drop events and trajectory deceleration features; Superimpose the event data obtained after preprocessing on the campus road network GIS layer in space, and generate a pothole damage heat map based on a preset damage determination rule; Extract high-risk road sections based on the pothole damage heat map, push the high-risk road section warning to the user, and output a dispatch work order to the maintenance department; The step of superimposing the event data obtained after preprocessing on the campus road network GIS layer in space and generating a pothole damage heat map based on a preset damage determination rule comprises: Map the spatial coordinates of the acceleration mutation events, tire pressure sudden drop events and trajectory deceleration features to the campus road network GIS layer, and construct a spatial aggregation model of event density distribution; Divide the campus road topology structure into dynamic adaptive grids, and perform weighted statistics on the cumulative event frequency in the grids, wherein the tire pressure sudden drop events are given the highest weight coefficient; Call the inherent facility position whitelist database on the road to filter acceleration mutation interference events caused by inherent facilities; Generate a damage probability heat map according to the event weight statistics result, and use a gradual color scale to represent the damage severity, wherein the area with high-frequency tire pressure events is rendered as the highest risk level; Integrate historical maintenance record data to attenuate and suppress the heat value of the repaired area, so as to realize the dynamic credibility evolution of the heat map; The step of calling the inherent facility position whitelist database on the road to filter acceleration mutation interference events caused by inherent facilities comprises: Construct a multi-vehicle type deceleration zone vibration feature library, extract the acceleration waveform features when passing through the deceleration zone at different speeds through machine learning, and the acceleration waveform features include amplitude envelope shape, frequency spectrum energy distribution and duration; When detecting an acceleration mutation event, calculate the dynamic time warping distance between its waveform and all templates in the feature library in real time, and if the minimum distance is lower than the adaptive matching threshold, it is determined as an inherent facility interference event; Combine the GIS coordinate data of the inherent facility and the vehicle passing direction to establish a three-dimensional space influence cone model, so that the filtering mechanism is triggered only when the acceleration mutation event meets the waveform matching and is located within the influence cone range; Continuously collect new acceleration mutation event data, and when the same type of events that do not match the feature library form a significant aggregation area after spatial clustering, automatically generate a suspected new deceleration zone learning sample and request manual annotation. 2.The campus entity trajectory analysis method of multi-source fused data according to claim 1, characterized in that, The intelligent riding terminal is integrated with a multi-modal sensor group, which includes a three-axis accelerometer for detecting vertical impact acceleration, a tire pressure sensor for monitoring tire pressure changes, a trajectory recording unit supporting GPS positioning, a solar power supply unit for power supply and a low-power communication unit, and the low-power communication unit only uploads data when the acceleration peak or the tire pressure change rate is greater than a set value. 3.The campus entity trajectory analysis method of multi-source fused data according to claim 1, characterized in that, The step of preprocessing the original data and extracting acceleration mutation events, tire pressure sudden drop events and trajectory deceleration features specifically comprises: The triaxial acceleration data is subjected to sliding window filtering processing to identify the instantaneous impact features in the vertical direction, and when the impact intensity exceeds the preset conventional road surface bump threshold, it is marked as an acceleration mutation event; Based on the tire pressure data, the tire pressure change trend is monitored in real time, and when a nonlinear cliff-like drop in the pressure value is detected, it is confirmed as a tire pressure sudden drop event; Based on the high-precision positioning trajectory data, the speed change rate of the continuous trajectory points is calculated, and when the vehicle speed suddenly and substantially attenuates and lasts for a specific period, it is extracted as a trajectory deceleration feature; A time-space correlation index of the three types of events is established, and a timestamp and a spatial confidence label are added to the event data to filter the coordinate abnormal values caused by satellite positioning drift. 4.The campus entity trajectory analysis method of multi-source fused data according to claim 1, characterized in that, The step of extracting high-risk road sections based on the pothole damage heat map and pushing high-risk road section warnings to users specifically comprises: The behavior data of the user is collected through the intelligent riding terminal, and the behavior data includes riding speed and route information; The high-risk road section distribution characteristics of the pothole damage heat map are matched with the behavior data of the user in terms of spatial correlation; When the user approaches the high-risk road section in the heat map, a prompt instruction is sent to the intelligent riding terminal, so that the intelligent riding terminal emits warning sound and light, and the intelligent riding terminal also integrates a sound and light alarm.

5. The campus subject trajectory analysis system of multi-source fused data, applied to the campus subject trajectory analysis method of multi-source fused data as claimed in any one of claims 1 to 4, characterized in that, The system comprises: A data synchronous acquisition module is configured to synchronously acquire original data of a vehicle through an intelligent riding terminal, the original data including triaxial acceleration data, tire pressure data and high-precision positioning trajectory data, and the intelligent riding terminal is suitable for authorized non-motor vehicles within a campus; An event extraction module is configured to preprocess the original data and extract acceleration mutation events, tire pressure sudden drop events and trajectory deceleration features; A heat map generation module is configured to perform spatial superposition on the event data obtained after preprocessing and a campus road network GIS layer, and generate a pothole damage heat map based on a preset damage determination rule; A linkage response module is configured to extract high-risk road sections based on the pothole damage heat map, push high-risk road section warnings to users, and output dispatch work orders to a maintenance department. 6.The campus principal entity trajectory analysis system of multi-source fused data according to claim 5, wherein, The event extraction module comprises: An acceleration filtering unit is configured to perform sliding window filtering processing on triaxial acceleration data to identify instantaneous impact features in the vertical direction, and when the impact intensity exceeds the preset conventional road surface bump threshold, it is marked as an acceleration mutation event; A tire pressure detection unit is configured to monitor the tire pressure change trend in real time based on the tire pressure data, and when a nonlinear cliff-like drop in the pressure value is detected, it is confirmed as a tire pressure sudden drop event; A speed attenuation extraction unit is configured to calculate the speed change rate of the continuous trajectory points based on the high-precision positioning trajectory data, and when the vehicle speed suddenly and substantially attenuates and lasts for a specific period, it is extracted as a trajectory deceleration feature; An event correlation unit is configured to establish a time-space correlation index of the three types of events, and add a timestamp and a spatial confidence label to the event data to filter the coordinate abnormal values caused by satellite positioning drift. 7.The campus entity trajectory analysis system of multi-source fused data according to claim 5, wherein, The heat map generation module comprises: An event space mapping unit is configured to map the spatial coordinates of the acceleration mutation event, the tire pressure drop event and the trajectory deceleration feature to a campus road network GIS layer, and construct a spatial aggregation model of event density distribution; A weighted statistical unit is configured to divide a dynamic adaptive grid according to a campus road topology structure, and perform weighted statistics on cumulative event frequencies in the grid, wherein the tire pressure drop event is given the highest weight coefficient; An interference filtering unit is configured to call an inherent facility position whitelist database on a road, and filter acceleration mutation interference events caused by inherent facilities; A thermal rendering output unit is configured to generate a damage probability thermal map according to the event weight statistical result, and use a gradient color scale to represent the damage severity, wherein a high-frequency tire pressure event area is rendered as the highest risk level; A dynamic attenuation suppression unit is configured to integrate historical repair record data, and attenuate and suppress the thermal values of repaired areas, so as to realize dynamic credibility evolution of the thermal map. 8.The campus entity trajectory analysis system of multi-source fused data according to claim 5, wherein, The linkage response module comprises: A behavior data acquisition unit is configured to acquire behavior data of a user through an intelligent riding terminal, wherein the behavior data comprises riding speed and route information; A thermal path matching unit is configured to perform spatial correlation matching based on the high-risk road segment distribution characteristics of the pothole damage thermal map and the behavior data of the user; An alarm triggering unit is configured to send a prompt instruction to the intelligent riding terminal when the user approaches a high-risk road segment in the thermal map, so that the intelligent riding terminal emits an alarm sound and light, and the intelligent riding terminal further integrates a sound and light alarm.

Images