A method, system and device for shared e-bike unmanned patrol
By using unmanned patrol methods for shared electric bicycles, real-time vehicle data is acquired, patrol intervals are dynamically adjusted, location drift is corrected, and electronic fences are optimized. This solves the problems of insufficient patrol coverage and misjudgment in the management of shared electric bicycles, and achieves efficient and accurate detection and management of illegal vehicles.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- 泰安市东信智联信息科技有限公司
- Filing Date
- 2026-02-05
- Publication Date
- 2026-06-23
AI Technical Summary
The current management of shared electric bicycles suffers from problems such as limited patrol coverage, high cost of manual patrols, large positioning errors, and difficulty in timely detection of illegal vehicles, which negatively impacts urban traffic order and the city's appearance.
By adopting an unmanned patrol method for shared electric bicycles, the patrol interval is dynamically adjusted by acquiring vehicle location data in real time, correcting positioning drift, optimizing electronic fence boundaries, and combining virtual boundary auxiliary points to achieve accurate violation judgment and uninterrupted patrol of vehicles.
It enables real-time and accurate detection of illegal vehicles throughout the city, reducing labor costs, minimizing misjudgments, and improving management efficiency and resource utilization.
Smart Images
Figure QLYQS_1 
Figure QLYQS_7
Abstract
Description
Technical Field
[0001] This invention belongs to the field of unmanned patrol technology for shared electric bicycles, and specifically relates to a method, system and equipment for unmanned patrol of shared electric bicycles. Background Technology
[0002] The shared electric bicycle industry in cities is currently experiencing rapid development, with a continuously expanding scale of vehicle deployment and a large number of operating entities. Due to the lack of unified industry management standards and cross-operator coordination mechanisms, each operator operates independently in vehicle dispatching and parking management, leading to frequent instances of haphazard parking of electric bicycles. This not only occupies public road resources such as urban motor vehicle lanes and sidewalks but also easily obstructs traffic, exacerbates regional road congestion, and negatively impacts urban traffic order and the city's appearance.
[0003] Existing technologies have limited coverage capabilities, and current management relies heavily on manual patrols. Due to limitations in labor costs and working hours, the patrol range cannot cover all areas of the city, and continuous patrols are not possible. There are blind spots in coverage in both time and space, making it difficult to detect and deal with illegally parked vehicles in a timely manner.
[0004] In existing compliant parking determination systems, when GPS signals drift or the density of electronic fence boundary points is low, the actual location of the vehicle is prone to deviate from the system's positioning coordinates, leading to increased boundary determination errors. This frequently results in compliant vehicles being misjudged as exceeding the fence and non-compliant vehicles not being identified, affecting management accuracy. Summary of the Invention
[0005] The purpose of this invention is to provide a method, system, and device for unmanned patrol of shared electric bicycles.
[0006] A method for unmanned patrol of shared electric bicycles includes the following steps:
[0007] S1. Real-time acquisition of location data and status of shared electric bicycles within a preset range, and calculation of vehicle density distribution based on location data;
[0008] S2. Based on vehicle density distribution, set differentiated dynamic patrol intervals for areas with different vehicle density distributions;
[0009] S3. Based on dynamic patrol, obtain the location coordinates of all shared electric bicycles within the preset range; if there is a shared electric bicycle that is not in use but the location coordinates are drifted, then call the preset spatial drift compensation process to correct the location coordinates that are drifting and obtain the corrected location coordinates.
[0010] S4. Use all corrected positioning coordinates and non-drifting positioning coordinates as the coordinates to be judged, and compare them with the optimized electronic fence boundary. If the coordinates to be judged do not fall within the optimized electronic fence boundary, the shared electric bicycle is judged to be illegally parked.
[0011] The optimized electronic fence boundary is obtained through the following operations:
[0012] For three consecutive vertices of the electronic fence polygon boundary, calculate the corner angle with the middle vertex as the vertex. If the absolute value of the difference between the corner angle and 180 degrees is greater than the preset angle threshold, it is determined that the corner is too curved. The two boundary line segments associated with the corner with excessive curvature are marked as line segments to be encrypted. On the line segments to be encrypted, at least one virtual boundary auxiliary point is inserted according to the length of the line segments to be encrypted and the corner angle. The electronic fence polygon boundary after inserting the virtual boundary auxiliary point is updated to the optimized electronic fence boundary.
[0013] S5. For shared electric bicycles determined to be illegally parked, obtain the location coordinates of the first threshold number of such bicycles. Each location coordinate is used as the location coordinate of the shared electric bicycle in S3. Execute S3 and S4. After all location coordinates have been executed in S3 and S4, if the number of times the bicycle is determined to be illegally parked is greater than the second threshold, the illegal parking determination is confirmed to be valid. Generate and push the illegal location of the illegally parked shared electric bicycle and the disposal instruction.
[0014] In S3, the preset spatial drift compensation process is called to correct the current positioning coordinates and obtain the corrected positioning coordinates. Specifically: S3.1, obtain the continuous positioning coordinates and signal quality of the shared electric bicycle whose positioning coordinates have drifted.
[0015] S3.2. In the continuous positioning coordinates, the result of subtracting the first point from the last point is taken as the dominant drift vector, and normalized to obtain the unit direction vector;
[0016] S3.3 From the continuous positioning coordinates, select the positioning coordinate with the best signal quality as the calibration point. If there are multiple positioning coordinates with the best signal quality, select the positioning coordinate with the earliest time as the calibration point.
[0017] S3.4 Calculate the drift prediction distance based on the confidence level of the calibration point and the dominant drift vector. Based on the drift prediction distance, the calibration point, and the unit direction vector, obtain the corrected coordinates.
[0018] The drift prediction distance is calculated based on the confidence level of the calibration point and the dominant drift vector, specifically as follows:
[0019] ,
[0020] in, As the dominant drift vector, This is an empirical attenuation factor, typically ranging from 0.3 to 0.6. For confidence level, For signal quality, This is a balancing term.
[0021] In S4, at least one virtual boundary auxiliary point is inserted based on the length and corner angle of the line segment to be encrypted, specifically:
[0022] Calculate the encryption priority coefficient P of the line segment to be encrypted.
[0023] ;
[0024] Where L is the length of the line segment to be encrypted. For the corner angle,
[0025] Determine the number of insertion points N.
[0026] N = ceil(P / S),
[0027] Where S is the length angle coefficient, and ceil is the round-up operation;
[0028] Divide the line segment to be encrypted into N+1 equal segments, and insert virtual boundary auxiliary points at the N division points within each segment.
[0029] The specific method for determining positioning drift in S3 is as follows: obtain at least three consecutive positioning coordinates of a stationary shared electric bicycle. If the three consecutive positioning coordinates move in the same direction, and the displacement distance between adjacent positioning coordinates satisfies a linear relationship, and the total displacement is greater than the error threshold, then positioning drift is determined to have occurred.
[0030] S2 sets differentiated dynamic patrol intervals for areas with different vehicle density distributions, specifically:
[0031] If the vehicle density is greater than the first density, the patrol interval is t1 seconds.
[0032] If the second density ≤ vehicle density ≤ first density, the patrol interval is t2 seconds;
[0033] If the vehicle density is less than the second density, the patrol interval is t3 seconds.
[0034] The t1 < t2 < t3.
[0035] In S5, the second threshold is half the number of the first threshold, and if it is a decimal, it is rounded up.
[0036] A shared electric bicycle unmanned patrol system, used to implement the aforementioned method for unmanned patrol of shared electric bicycles, includes:
[0037] The data acquisition module acquires the location data and status of shared electric bicycles within a preset range in real time, and calculates the vehicle density distribution based on the location data;
[0038] The dynamic patrol module sets differentiated dynamic patrol intervals for areas with different vehicle density distributions based on vehicle density distribution.
[0039] The coordinate correction module obtains the location coordinates of all shared electric bicycles within a preset range based on dynamic inspection. If a shared electric bicycle is not in use but its location coordinates have drifted, the preset spatial drift compensation process is called to correct the location coordinates that have drifted, thus obtaining the corrected location coordinates.
[0040] The violation determination module uses all corrected location coordinates and non-drift location coordinates as the coordinates to be determined, and compares them with the optimized electronic fence boundary. If the coordinates to be determined do not fall within the optimized electronic fence boundary, the shared electric bicycle is determined to be illegally parked.
[0041] The optimized electronic fence boundary is obtained through the following operations:
[0042] For three consecutive vertices of the electronic fence polygon boundary, calculate the corner angle with the middle vertex as the vertex. If the absolute value of the difference between the corner angle and 180 degrees is greater than the preset angle threshold, it is determined that the corner is too curved. The two boundary line segments associated with the corner with excessive curvature are marked as line segments to be encrypted. On the line segments to be encrypted, at least one virtual boundary auxiliary point is inserted according to the length of the line segments to be encrypted and the corner angle. The electronic fence polygon boundary after inserting the virtual boundary auxiliary point is updated to the optimized electronic fence boundary.
[0043] The violation review module obtains the location coordinates of a first threshold number for shared electric bicycles that are determined to be illegally parked. Each location coordinate is used as the location coordinate of the shared electric bicycle in the coordinate correction module. The coordinate correction module and the violation judgment module are executed. After all location coordinates have been executed, if the number of times the illegal parking is determined is greater than the second threshold, the illegal parking judgment is confirmed to be valid, and the illegal parking location of the shared electric bicycle and the disposal instruction are generated and pushed.
[0044] The second threshold is half the number of the first threshold; if it is a decimal, it is rounded up.
[0045] A device for unmanned patrol of shared electric bicycles includes a processor and a memory, wherein the processor executes a computer program stored in the memory to implement the method for unmanned patrol of shared electric bicycles.
[0046] The beneficial effects of this application are as follows:
[0047] This application enables unmanned patrols to detect violations by shared electric bicycles. It has a large coverage area, allows for real-time and uninterrupted patrols, reduces labor costs, minimizes blind spots, and promptly detects illegal vehicles.
[0048] By inserting auxiliary points into the line segment to be encrypted, the long line segment is divided into multiple short line segments. This makes the optimized electronic fence boundary more closely match the real geographical curve, greatly reducing model error and thus directly reducing misjudgment.
[0049] Correcting drift can eliminate errors caused by signal quality, improve the accuracy of judgment results, and increase the accuracy of shared electric bicycle positioning. Managers can then clearly understand the distribution of vehicles and violations throughout the city.
[0050] This application also makes dynamic planning for dynamic inspection intervals, which helps to rationally allocate computing power and improve resource utilization. Detailed Implementation
[0051] To further understand the content of this invention, the invention will be described in detail with reference to the embodiments.
[0052] This invention relates to a method for unmanned patrol of shared electric bicycles, which includes the following steps:
[0053] S1. Real-time acquisition of location data and status of shared electric bicycles within a preset range, and calculation of vehicle density distribution based on location data.
[0054] At a set base frequency, the location data and vehicle status of all online shared electric bicycles are obtained in real time. The vehicle status here refers to whether the shared electric bicycle is being ridden or locked.
[0055] Based on location data, the number of shared electric bicycles in a two-dimensional geographic grid within the target area is calculated, generating a vehicle distribution heatmap. This heatmap visually reflects areas of vehicle aggregation and sparseness. Based on the vehicle distribution heatmap, the vehicle density distribution is obtained, with the unit being vehicles / 100 square meters.
[0056] S2. Based on vehicle density distribution, set differentiated dynamic patrol intervals for areas with different vehicle density distributions.
[0057] Specifically, if the vehicle density is greater than the first density, the patrol interval is t1 seconds;
[0058] If the second density ≤ vehicle density ≤ first density, the patrol interval is t2 seconds;
[0059] If the vehicle density is less than the second density, the patrol interval is t3 seconds.
[0060] The t1 < t2 < t3.
[0061] Furthermore, the base frequency mentioned in S1 can be set to 30 seconds / time. If the identified area has a vehicle density of >5 vehicles / 100㎡, such as a subway entrance or a commercial center, the risk of violations is high and changes rapidly in such areas, so the patrol interval is shortened to 10 seconds. This means that the location data of vehicles in these areas will be obtained more frequently. Low-priority areas have a vehicle density of <1 vehicle / 100㎡, such as suburbs. These areas change slowly, so the patrol interval is extended to 60 seconds to save resources. Ordinary areas are automatically assigned to areas with densities between the above two, and the base frequency is used with a patrol interval set to 30 seconds.
[0062] Different patrol interval settings can allocate computing power to areas with higher density and reduce computing power allocation to areas with lower density.
[0063] S3. Based on dynamic inspection, obtain the location coordinates of all shared electric bicycles within the preset range; if there is a shared electric bicycle that is not in use but the location coordinates are drifted, then call the preset spatial drift compensation process to correct the location coordinates that are drifting and obtain the corrected location coordinates.
[0064] Furthermore, the following steps in this application are mainly for shared electric bicycles in an unused state. Shared electric bicycles in use are not applicable to this application. In addition, shared electric bicycles that are temporarily parked while in use are also not applicable to this application.
[0065] The shared electric bicycles are dynamically inspected according to the dynamic inspection interval to obtain their location coordinates. It is necessary to determine whether the location coordinates of the shared electric bicycles in the unused state have drifted. At least three consecutive location coordinates of the shared electric bicycles in the stationary state are obtained. If the movement direction of the three consecutive location coordinates is consistent, the displacement distance between adjacent location coordinates satisfies the linear relationship, and the total displacement is greater than the error threshold, then it is determined that location drift has occurred.
[0066] The error threshold is dynamically adjusted based on the signal quality of the positioning coordinates; the lower the signal quality, the larger the error threshold.
[0067] The preset spatial drift compensation process is invoked to correct the current positioning coordinates where positioning drift has occurred, as follows:
[0068] S3.1 Obtain the continuous positioning coordinates and signal quality of the shared electric bicycles whose positioning coordinates have drifted.
[0069] S3.2. In the continuous positioning coordinates, the result of subtracting the first point from the last point is taken as the dominant drift vector. Normalize it to obtain the unit direction vector, which represents the overall drift direction from the first point to the last point.
[0070] S3.3 From the continuous positioning coordinates, select the positioning coordinate with the best signal quality as the calibration point. If there are multiple positioning coordinates with the best signal quality, select the positioning coordinate with the earliest time as the calibration point. In a static state, the earliest observation point with the best signal quality is most likely to be close to the true position. This is because drift is usually caused by accumulation or sudden jumps, and the starting point has the weakest relative interference.
[0071] S3.4 Calculate the drift prediction distance based on the confidence level of the calibration point and the dominant drift vector. Based on the drift prediction distance, the calibration point, and the unit direction vector, obtain the corrected coordinates.
[0072] Specifically, calculate the drift prediction distance:
[0073] ,
[0074] in, As the dominant drift vector, This is an empirical attenuation factor, typically ranging from 0.3 to 0.6. For confidence level, For signal quality, This is a balancing term.
[0075] After obtaining the drift prediction distance, move the drift prediction distance in the direction of the unit direction vector of the calibration point to obtain the corrected coordinates.
[0076] S4. Use all corrected positioning coordinates and non-drifting positioning coordinates as the coordinates to be judged, and compare them with the optimized electronic fence boundary. If the coordinates to be judged do not fall within the optimized electronic fence boundary, the shared electric bicycle is judged to be illegally parked.
[0077] Obtain the polygonal boundary of the electronic fence, optimize the polygonal boundary to obtain the optimized electronic fence boundary. The optimized electronic fence boundary is obtained through the following operations:
[0078] For three consecutive vertices of the electronic fence polygon boundary, calculate the corner angle with the middle vertex as the vertex. If the absolute value of the difference between the corner angle and 180 degrees is greater than the preset angle threshold, the corner is determined to be too curved. The two boundary line segments associated with the corner that is too curved are marked as line segments to be encrypted. On the line segments to be encrypted, at least one virtual boundary auxiliary point is inserted according to the length of the line segments to be encrypted and the corner angle. The electronic fence polygon boundary after inserting the virtual boundary auxiliary point is updated to the optimized electronic fence boundary.
[0079] Preferably, at least one virtual boundary auxiliary point is inserted according to the length and corner angle of the line segment to be encrypted, specifically:
[0080] Calculate the encryption priority coefficient P of the line segment to be encrypted.
[0081] ;
[0082] Where L is the length of the line segment to be encrypted. For corner angles, the absolute value design can handle concave or convex corners.
[0083] Determine the number of insertion points N.
[0084] N = ceil(P / S),
[0085] Where S is the length angle coefficient, and ceil is the round-up operation;
[0086] Divide the line segment to be encrypted into N+1 equal segments, and insert virtual boundary auxiliary points at the N division points within each segment.
[0087] Once the optimized electronic fence boundary is obtained, the corrected positioning coordinates and the positioning coordinates that have not drifted are used as the coordinates to be judged and compared with the optimized electronic fence boundary. If the coordinates to be judged do not fall within the optimized electronic fence boundary, the shared electric bicycle is judged to be illegally parked.
[0088] If the coordinates to be determined fall within the optimized electronic fence boundary, the shared electric bicycle is deemed to be parked compliantly. No further processing is performed.
[0089] The goal of this application in optimizing the electronic fence boundary is to segment significant inflection points, thereby reducing the possibility of errors in subsequent judgments and improving the accuracy of judgments.
[0090] S5. For shared electric bicycles determined to be illegally parked, obtain the location coordinates of the first threshold number of such bicycles. Each location coordinate is used as the location coordinate of the shared electric bicycle in S3. Execute S3 and S4. After all location coordinates have been executed in S3 and S4, if the number of times the bicycle is determined to be illegally parked is greater than the second threshold, the illegal parking determination is confirmed to be valid. Generate and push the illegal location of the illegally parked shared electric bicycle and the disposal instruction.
[0091] The accuracy of the violation determination needs to be verified. Further, according to the patrol interval, the location coordinates of the shared electric bicycles determined to be illegally parked are obtained, and a first threshold number of location coordinates are obtained. For each coordinate, S3 and S4 are executed. After execution, if the number of times the illegal parking is determined is greater than the second threshold, the illegal parking determination is confirmed to be valid.
[0092] If the number of times a vehicle is judged to be illegally parked is less than or equal to the second threshold number, it means that the violation judgment of S4 is incorrect, and the current shared electric bicycle will not be dealt with temporarily.
[0093] Furthermore, for areas with different vehicle density distributions, the first threshold number is set dynamically. The first threshold number for areas with different vehicle density distributions can be set to a uniform total duration. Within this total duration, the specific value of the first threshold number is set according to the different dynamic patrol intervals.
[0094] The second threshold can be half the number of the first threshold, with decimals rounded up to ensure accurate judgment. Then, the location of the illegally parked shared electric bicycle and the disposal instruction are generated and pushed. If the number of violations is less than or equal to the number of the second threshold, the location coordinates of the shared electric bicycle at the time of the first judgment may be abnormal. After repeated judgments by S5, it is not considered illegally parked and no action is taken.
[0095] This application S5 uses verification to avoid misjudgments caused by single data errors, thereby improving the accuracy of patrolling illegal shared electric bicycles.
[0096] A shared electric bicycle unmanned patrol system, used to implement the aforementioned method for unmanned patrol of shared electric bicycles, includes:
[0097] The data acquisition module acquires the location data and status of shared electric bicycles within a preset range in real time, and calculates the vehicle density distribution based on the location data;
[0098] The dynamic patrol module sets differentiated dynamic patrol intervals for areas with different vehicle density distributions based on vehicle density distribution.
[0099] The coordinate correction module obtains the location coordinates of all shared electric bicycles within a preset range based on dynamic inspection. If a shared electric bicycle is not in use but its location coordinates have drifted, the preset spatial drift compensation process is called to correct the location coordinates that have drifted, thus obtaining the corrected location coordinates.
[0100] The violation determination module uses all corrected location coordinates and non-drift location coordinates as the coordinates to be determined, and compares them with the optimized electronic fence boundary. If the coordinates to be determined do not fall within the optimized electronic fence boundary, the shared electric bicycle is determined to be illegally parked.
[0101] The optimized electronic fence boundary is obtained through the following operations:
[0102] For three consecutive vertices of the electronic fence polygon boundary, calculate the corner angle with the middle vertex as the vertex. If the absolute value of the difference between the corner angle and 180 degrees is greater than the preset angle threshold, it is determined that the corner is too curved. The two boundary line segments associated with the corner with excessive curvature are marked as line segments to be encrypted. On the line segments to be encrypted, at least one virtual boundary auxiliary point is inserted according to the length of the line segments to be encrypted and the corner angle. The electronic fence polygon boundary after inserting the virtual boundary auxiliary point is updated to the optimized electronic fence boundary.
[0103] The violation review module obtains the location coordinates of a first threshold number for shared electric bicycles that are determined to be illegally parked. Each location coordinate is used as the location coordinate of the shared electric bicycle in the coordinate correction module. The coordinate correction module and the violation judgment module are executed. After all location coordinates have been executed, if the number of times the illegal parking is determined is greater than the second threshold, the illegal parking judgment is confirmed to be valid, and the illegal parking location of the shared electric bicycle and the disposal instruction are generated and pushed.
[0104] The second threshold is half the number of the first threshold; if it is a decimal, it is rounded up.
[0105] A device for unmanned patrol of shared electric bicycles includes a processor and a memory, wherein the processor executes a computer program stored in the memory to implement the method for unmanned patrol of shared electric bicycles.
Claims
1. A method for unmanned patrol of shared electric bicycles, characterized in that, Includes the following steps: S1. Real-time acquisition of location data and status of shared electric bicycles within a preset range, and calculation of vehicle density distribution based on location data; S2. Based on vehicle density distribution, set differentiated dynamic patrol intervals for areas with different vehicle density distributions; Specifically: If the vehicle density is greater than the first density, the patrol interval is t1 seconds. If the second density ≤ vehicle density ≤ first density, the patrol interval is t2 seconds; If the vehicle density is less than the second density, the patrol interval is t3 seconds. The t1 < t2 < t3; S3. Based on dynamic patrol, obtain the location coordinates of all shared electric bicycles within the preset range; if there is a shared electric bicycle that is not in use but the location coordinates are drifted, then call the preset spatial drift compensation process to correct the location coordinates that are drifting and obtain the corrected location coordinates. The specific method for determining location drift is as follows: obtain at least three consecutive location coordinates of a stationary shared electric bicycle. If the movement direction of the three consecutive location coordinates is consistent, the displacement distance between adjacent location coordinates satisfies a linear relationship, and the total displacement is greater than the error threshold, then it is determined that location drift has occurred. S3.1 Obtain the continuous positioning coordinates and signal quality of the shared electric bicycles whose positioning coordinates have drifted; S3.
2. In the continuous positioning coordinates, the result of subtracting the first point from the last point is taken as the dominant drift vector, and normalized to obtain the unit direction vector; S3.3 From the continuous positioning coordinates, select the positioning coordinate with the best signal quality as the calibration point. If there are multiple positioning coordinates with the best signal quality, select the positioning coordinate with the earliest time as the calibration point. S3.4 Calculate the drift prediction distance based on the confidence level of the calibration point and the dominant drift vector. Based on the drift prediction distance, calibration point, and unit direction vector, obtain the corrected coordinates. The drift prediction distance is calculated based on the confidence level of the calibration point and the dominant drift vector, specifically as follows: ， in, As the dominant drift vector, This is an empirical attenuation factor, typically ranging from 0.3 to 0.
6. For confidence level, For signal quality, It is a balancing term; After obtaining the drift prediction distance, move the drift prediction distance in the direction of the unit direction vector of the calibration point to obtain the corrected coordinates; S4. Use all corrected positioning coordinates and non-drifting positioning coordinates as the coordinates to be judged, and compare them with the optimized electronic fence boundary. If the coordinates to be judged do not fall within the optimized electronic fence boundary, the shared electric bicycle is judged to be illegally parked. The optimized electronic fence boundary is obtained through the following operations: For three consecutive vertices of the electronic fence polygon boundary, calculate the corner angle with the middle vertex as the vertex. If the absolute value of the difference between the corner angle and 180 degrees is greater than the preset angle threshold, it is determined that the corner is too curved. The two boundary line segments associated with the corner with excessive curvature are marked as line segments to be encrypted. On the line segments to be encrypted, at least one virtual boundary auxiliary point is inserted according to the length of the line segments to be encrypted and the corner angle. The electronic fence polygon boundary after inserting the virtual boundary auxiliary point is updated to the optimized electronic fence boundary. Based on the length and corner angle of the line segment to be encrypted, insert at least one virtual boundary auxiliary point, specifically: Calculate the encryption priority coefficient P of the line segment to be encrypted. ； Where L is the length of the line segment to be encrypted. For the corner angle, Determine the number of insertion points N. N = ceil(P / S), Where S is the length angle coefficient, and ceil is the round-up operation; Divide the line segment to be encrypted into N+1 equal segments, and insert virtual boundary auxiliary points at the N equal division points inside it; S5. For shared electric bicycles determined to be illegally parked, obtain the location coordinates of the first threshold number of such bicycles. Each location coordinate is used as the location coordinate of the shared electric bicycle in S3. Execute S3 and S4. After all location coordinates have been executed in S3 and S4, if the number of times the bicycle is determined to be illegally parked is greater than the second threshold, the illegal parking determination is confirmed to be valid. Generate and push the illegal location of the illegally parked shared electric bicycle and the disposal instruction.
2. The method for unmanned patrol of shared electric bicycles according to claim 1, characterized in that, In S5, the second threshold is half the number of the first threshold, and if it is a decimal, it is rounded up.
3. A shared electric bicycle unmanned patrol system, used to implement the method for unmanned patrol of shared electric bicycles as described in any one of claims 1-2, characterized in that, include: The data acquisition module acquires the location data and status of shared electric bicycles within a preset range in real time, and calculates the vehicle density distribution based on the location data; The dynamic patrol module sets differentiated dynamic patrol intervals for areas with different vehicle density distributions based on vehicle density distribution. The coordinate correction module obtains the location coordinates of all shared electric bicycles within a preset range based on dynamic inspection. If a shared electric bicycle is not in use but its location coordinates have drifted, the preset spatial drift compensation process is called to correct the location coordinates that have drifted, thus obtaining the corrected location coordinates. The violation determination module uses all corrected location coordinates and non-drift location coordinates as the coordinates to be determined, and compares them with the optimized electronic fence boundary. If the coordinates to be determined do not fall within the optimized electronic fence boundary, the shared electric bicycle is determined to be illegally parked. The optimized electronic fence boundary is obtained through the following operations: For three consecutive vertices of the electronic fence polygon boundary, calculate the corner angle with the middle vertex as the vertex. If the absolute value of the difference between the corner angle and 180 degrees is greater than the preset angle threshold, it is determined that the corner is too curved. The two boundary line segments associated with the corner with excessive curvature are marked as line segments to be encrypted. On the line segments to be encrypted, at least one virtual boundary auxiliary point is inserted according to the length of the line segments to be encrypted and the corner angle. The electronic fence polygon boundary after inserting the virtual boundary auxiliary point is updated to the optimized electronic fence boundary. The violation review module obtains the location coordinates of a first threshold number for shared electric bicycles that are determined to be illegally parked. Each location coordinate is used as the location coordinate of the shared electric bicycle in the coordinate correction module. The coordinate correction module and the violation judgment module are executed. After all location coordinates have been executed, if the number of times the illegal parking is determined is greater than the second threshold, the illegal parking judgment is confirmed to be valid, and the illegal parking location of the shared electric bicycle and the disposal instruction are generated and pushed.
4. The system for unmanned patrol of shared electric bicycles according to claim 3, characterized in that, The second threshold is half the number of the first threshold; if it is a decimal, it is rounded up.
5. A device for unmanned patrol of shared electric bicycles, characterized in that, It includes a processor and a memory, wherein the processor executes a computer program stored in the memory to implement a method for unmanned patrol of shared electric bicycles as described in any one of claims 1-2.