Shared electric bicycle fixed-point directional parking system

Abstract

The invention discloses a shared electric bicycle fixed-point directional parking system, and relates to the technical field of fixed-point directional parking, a GPS signal quality detection mechanism is adopted, the variance of GPS data is calculated through multiple times of sampling, the positioning precision is evaluated, if the signal is stable, GPS positioning is directly adopted, otherwise, UWB is called for position correction, and the positioning accuracy is improved. The final position information is calculated by adopting a weighted fusion algorithm, so that the stability of vehicle positioning is improved; a polygon bounding algorithm and Euclidean distance calculation are adopted for matching of the parking area, it is guaranteed that the parking position of the vehicle meets the standard, and misjudgment possibly caused by positioning errors is effectively filtered out by setting the tolerance range of the parking area; in the aspect of directional parking, an ultra-wide-angle camera is installed in a parking area, visual recognition is conducted on a parked vehicle in combination with an edge calculation unit, the main direction angle of the vehicle is calculated through a minimum enclosing rectangle method, and meanwhile it is guaranteed that the vehicle is parked in the preset direction by combining an IMU sensor and an electronic compass and calculating attitude information and a magnetic azimuth angle.

Description

Technical Field

[0001] This invention relates to the field of fixed-point and directional parking technology, and in particular to a fixed-point and directional parking system for shared electric bicycles. Background Technology

[0002] The widespread use of shared electric bicycles has greatly facilitated short-distance travel in cities. To regulate management and ensure smooth road traffic, fixed-point and directional parking is now commonly used, guiding users to park in designated areas through technologies such as electronic fences, QR code recognition, and smart locks.

[0003] Shared electric bicycle parking spots are mostly located in transportation hubs, commercial areas, residential areas, and public transportation interchanges. These areas typically have high pedestrian traffic and parking demand, but limited parking space. Therefore, more precise technological means are needed to ensure that vehicles are parked in compliant areas. However, current parking area management relies heavily on GPS positioning, which is easily affected by environmental factors. In environments such as overpasses and interchanges, GPS signal stability is poor, and the system is prone to errors in determining parking locations, resulting in a poor user experience.

[0004] To address the issue of unstable GPS signals, some traditional solutions involve expanding the electronic fence area or adding signal base stations to reduce misjudgments through fault tolerance mechanisms. However, excessively large fence areas can lead to inaccurate parking by users, affecting urban management. While adding base stations or optimizing signal receiving equipment can improve positioning accuracy in some areas, GPS signal drift cannot be completely avoided in environments such as overpasses and under viaducts. Therefore, improving positioning accuracy in these special environments has become an urgent problem to be solved for the fixed-point and directional parking of shared electric bicycles. Summary of the Invention

[0005] In view of the aforementioned existing problems, the present invention is proposed.

[0006] This invention provides a shared electric bicycle fixed-point and directional parking system that solves the problems of GPS signal interference leading to misjudgment of electronic fences, expanding the fence to reduce management accuracy, and signal optimization making it difficult to completely eliminate drift problems in elevated areas.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0008] This invention provides a fixed-point and directional parking system for shared electric bicycles, comprising:

[0009] The vehicle positioning module, including an ultra-wideband (UWB) positioning unit and a GPS signal quality detection unit, is used to obtain the current location of the shared electric bicycle.

[0010] The parking area matching module is used to determine whether a vehicle is located within a legal parking area;

[0011] The visual recognition module includes an ultra-wide-angle camera and an edge computing unit, which is used to perform visual recognition of parked vehicles in the parking area based on the ultra-wide-angle camera and the edge computing unit to determine whether the vehicle is parked in a preset direction, and to verify the direction data recognized by the user terminal.

[0012] The vehicle orientation detection module, including an inertial measurement unit (IMU) and an electronic compass, is used to acquire vehicle parking orientation data and compare it with the preset orientation standard of the parking area.

[0013] The vehicle parking verification module performs a secondary verification of the vehicle's parking posture based on the visual recognition module after the user has finished parking. It also uses the edge computing unit to automatically determine the vehicle's parking position and angle, and calls up the electronic fence data to confirm again whether the vehicle's position exceeds the compliant parking area.

[0014] The parking management module is used to unlock the settlement function after confirming that the vehicle's location and parking direction are compliant, allowing users to complete the vehicle locking operation and synchronizing the vehicle's parking location, parking direction, and parking time data to the cloud database.

[0015] As a preferred embodiment of the shared electric bicycle fixed-point and directional parking system described in this invention, the vehicle positioning module detects the quality of the Global Navigation Satellite System (GPS) signal. When the GPS signal meets the preset stability requirements, GPS data is used for positioning. If the GPS signal quality is lower than the preset threshold, the Ultra Wideband (UWB) positioning unit is called for position correction.

[0016] The GPS signal quality detection unit is used to determine whether the GPS signal meets the preset stability requirements, and when the GPS signal quality is lower than the threshold, it calls the ultra-wideband (UWB) positioning unit to perform position correction.

[0017] As a preferred embodiment of the shared electric bicycle fixed-point and directional parking system of the present invention, the parking area matching module, based on the vehicle location information obtained by the vehicle positioning module, calls the parking area database, compares the current vehicle location with the boundary data of the preset parking area, and determines whether the vehicle is within the compliant parking range.

[0018] The ultra-wide-angle camera in the visual recognition module is mounted on a fixed bracket or light pole within the parking area, covering the entire parking area, and the image data is processed locally by the edge computing unit.

[0019] As a preferred embodiment of the shared electric bicycle fixed-point and directional parking system of the present invention, wherein: the IMU sensor and electronic compass in the vehicle orientation detection module are integrated into the shared electric bicycle to obtain the vehicle's posture information and parking angle in real time;

[0020] In the vehicle orientation detection module, if the vehicle's position is compliant but its direction deviates from the preset direction, a prompt message to adjust the direction is sent to the user, and the sending method can be the user terminal or the vehicle's voice system.

[0021] As a preferred embodiment of the shared electric bicycle fixed-point and directional parking system of the present invention, the vehicle parking verification module includes a parking area monitoring unit based on AI vision, which is used to identify whether the parked vehicle occupies the sidewalk, blind path or non-parking area, and to send alarm information to the parking management module when illegal parking occurs.

[0022] As a preferred embodiment of the shared electric bicycle fixed-point and directional parking system described in this invention, the parking method of the system is as follows:

[0023] Step S1: After the user rides to the target parking area, the vehicle positioning module obtains the current vehicle location information and detects the GPS signal quality.

[0024] If the GPS signal quality meets the preset stability requirements, then GPS data will be used for vehicle positioning.

[0025] If the GPS signal quality is lower than a preset threshold, the ultra-wideband (UWB) positioning unit will be invoked to correct the vehicle's position.

[0026] Step S2: Based on the vehicle location information determined in step S1, call the parking area database, compare the current vehicle location with the preset parking area boundary data, and determine whether the current vehicle is within the compliant parking area.

[0027] Step S3: Within the parking area, visual recognition of parked vehicles is performed based on an ultra-wide-angle camera and an edge computing unit to determine whether the vehicles are parked in a preset direction.

[0028] Vehicle parking direction data is acquired using an onboard inertial measurement unit (IMU) and an electronic compass, and compared with the preset direction standard of the parking area.

[0029] If the vehicle is in the correct position but deviates from the preset direction, a prompt message will be sent to the user to adjust the direction.

[0030] Step S4: After the user finishes parking, the vehicle's parking posture is checked a second time based on the visual recognition module, and the parking position and angle of the vehicle are automatically determined by the edge computing unit.

[0031] In step S3, based on the vehicle location information determined in step S1, and by calling the electronic fence data, it is confirmed again whether the vehicle location exceeds the compliant parking area.

[0032] In step S5, after the parking management module confirms that the vehicle location and parking direction are compliant, it unlocks the settlement function, allowing the user to complete the vehicle locking operation, and synchronizes data such as vehicle parking location, parking direction, and parking time to the cloud database.

[0033] In a preferred embodiment of the shared electric bicycle fixed-point and directional parking system described in this invention, the step of obtaining the current vehicle location information by the vehicle positioning module and simultaneously detecting the GPS signal quality is as follows:

[0034] The vehicle positioning module obtains the current vehicle's GPS data through the GPS signal quality detection unit, which is represented as:

[0035] P GPS =(x GPS ,y GPS ,z GPS ,t GPS SNR GPS ),

[0036] Among them, P GPS This represents the raw location information collected by GPS, x GPS This represents the coordinates of the vehicle's position provided by GPS along the X-axis, y-axis. GPS This represents the coordinates of the vehicle's position along the Y-axis provided by GPS, z. GPS t represents the coordinate value of the vehicle's position in the Z-axis direction provided by GPS. GPS SNR represents the timestamp of a GPS signal. GPS Indicates the signal-to-noise ratio of GPS signals.

[0037] Multiple GPS data collections were performed to calculate the position deviation variance. The calculation formula is:

[0038]

[0039] in, The variance of the GPS location data is represented by N, where N represents the number of GPS data points sampled. The deviation of the j-th GPS data point from the mean is represented by the following formula:

[0040]

[0041] in, These represent the coordinates of the j-th GPS sampling point. These represent the mean coordinates of all GPS sampling points. The average offset distance of the GPS sampling points is expressed by the following formula:

[0042]

[0043] Determine whether the GPS signal quality meets the preset stability requirements, and set the GPS signal quality stability threshold τ. GPS If the following conditions are met, the GPS signal quality is considered stable, and GPS data is used for positioning; otherwise, UWB is used for position correction:

[0044]

[0045] Where, τ GPS SNR is the variance threshold for GPS signal quality. min The minimum acceptable signal-to-noise ratio threshold for GPS signals.

[0046] If the GPS signal quality is below the threshold, UWB will be used for positioning. The UWB location calculation is as follows:

[0047] P UWB =(x UWB ,y UWB ,z UWB ,t UWB ),

[0048] Among them, P UWB Indicates the location information collected by UWB, x UWB ,y UWB ,z UWB Represent the vehicle position coordinates calculated by UWB, t UWB Indicates the UWB collection timestamp.

[0049] The final vehicle position is calculated using a weighted fusion method, and the fusion formula is as follows:

[0050] P = w GPS P GPS +w UWB P UWB ,

[0051] Where P represents the final fused vehicle location information, and w GPS ,w UWB The weights for GPS and UWB are calculated as follows:

[0052]

[0053] Among them, SNR UWB The signal-to-noise ratio of the UWB signal.

[0054] As a preferred embodiment of the shared electric bicycle fixed-point and directional parking system of the present invention, the step of comparing the current vehicle position with the preset parking area boundary data to determine whether the current vehicle is within the compliant parking range is as follows:

[0055] Access the parking area database to obtain the boundary data of compliant parking areas. Assume the parking area database provides the boundary information of compliant parking areas.

[0056] R = {(x k ,y k |k=1,2,…,M},

[0057] Where R is the set of boundary points of the compliant parking area, M represents the number of boundary points, (x k ,y k () represents the coordinates of the boundary point.

[0058] An algorithm is used to determine whether a vehicle's position P(x,y) is within a compliant parking area by identifying points inside a polygon. The number of times the boundary is crossed is calculated using the following formula:

[0059]

[0060] Where C represents the number of times the vehicle crosses the boundary. If C is odd, the vehicle is within the compliant parking area; otherwise, the vehicle is outside the compliant area.

[0061] The minimum Euclidean distance between the vehicle's current position and the boundary point of the parking area is calculated using the following formula:

[0062]

[0063] Where, d min This indicates the shortest distance from the vehicle's location to the boundary of the compliant parking area.

[0064] Let the tolerance range of the parking area be δ. R If the following conditions are met:

[0065] d min ≤δ R If the vehicle is parked within the permitted parking area, it is considered to be in a compliant parking area; otherwise, the vehicle is parked improperly.

[0066] As a preferred embodiment of the shared electric bicycle fixed-point and directional parking system described in this invention, the step of visually recognizing parked vehicles based on an ultra-wide-angle camera and an edge computing unit to determine whether the vehicles are parked in a preset direction is as follows:

[0067] The ultra-wide-angle camera in the visual recognition module acquires image data of the parking area, represented as follows:

[0068] I = {I f |f=1,2,…,F},

[0069] Where I is the image dataset of the parking area, I f This represents the f-th frame of the image, where F is the total number of image frames.

[0070] The edge computing unit is used to extract the contour of the image data, specifically the contour region of the parked vehicle. The extraction formula is as follows:

[0071] C V ={(x c ,y c |c=1,2,…,C},

[0072] Among them, C V The set of pixels representing the vehicle outline, where C is the total number of outline pixels, (x c ,y c () represents the pixel coordinates of the c-th contour point.

[0073] The vehicle's principal steering angle is calculated using the minimum bounding rectangle method. The formula is as follows:

[0074]

[0075] Where, θ V The vehicle heading angle for visual recognition, x max ,x min The coordinates of the rightmost and leftmost points of the contour boundary, y max ,y min The coordinates of the uppermost and lowermost points of the contour boundary.

[0076] The vehicle orientation detection module acquires vehicle attitude information P through an IMU sensor. IMU :

[0077] P IMU =(α IMU ,β IMU ,γ IMU ,t IMU ),

[0078] Among them, P IMU α represents the attitude information acquired by the IMU. IMU ,β IMU ,γ IMU These represent the vehicle's roll angle, pitch angle, and yaw angle, respectively. IMU Indicates the timestamp of IMU data.

[0079] The electronic compass acquires magnetic field data and calculates the vehicle's magnetic azimuth angle using the following formula:

[0080]

[0081] Where, θ M B represents the vehicle heading angle calculated by the electronic compass. x B y These represent the magnetic induction intensity of the magnetic field sensor along the X and Y axes, respectively.

[0082] Compare whether the vehicle's steering angle conforms to the preset direction. Let θ be the preset direction standard for the parking area. std Determine the vehicle's directional deviation:

[0083] Δθ=min(|θ V -θ std |,|θ M -θ std |),

[0084] Where Δθ represents the angle of deviation between the vehicle's direction and the preset direction.

[0085] If Δθ exceeds the threshold θ max The user terminal and the vehicle's voice module send adjustment prompts.

[0086] As a preferred embodiment of the shared electric bicycle fixed-point and directional parking system described in this invention, the step of performing a secondary verification of the vehicle's parking posture based on a visual recognition module, and automatically determining the vehicle's parking position and angle using an edge computing unit, is as follows:

[0087] After the user completes parking, the visual recognition module re-captures the parking area image I':

[0088] I'={I' f |f=1,2,…,F'},

[0089] Where I' represents the image dataset during the parking review phase, I' f This represents the f-th frame of the verification image, where F' is the total number of verification image frames.

[0090] The vehicle outline is identified again, and the vehicle outline for the review stage is calculated using the following formula:

[0091] C' V ={(x' c ,y' c |c=1,2,…,C'},

[0092] Among them, C' V Let C' be the set of pixels representing the vehicle outline during the verification phase, and let C' be the total number of outline pixels during the verification phase.

[0093] The formula for calculating the vehicle steering angle during the second verification stage is as follows:

[0094]

[0095] Where, θ' V x' represents the vehicle orientation angle for visual recognition during the second verification stage. max ,x' minDuring the review phase, the coordinates of the rightmost and leftmost ends of the vehicle outline boundary, y' max ,y' min The coordinates of the top and bottom edges of the vehicle's outline boundary during the verification phase.

[0096] Accessing the electronic fence data confirms the location's compliance; the electronic fence database provides boundary information for the parking area. E :

[0097] R E ={(x m ,y m )|m=1,2,…,M'},

[0098] Among them, R E M' is the set of boundary points for the parking area defined by the electronic fence, where M' is the number of electronic fence boundary points. m ,y m ( ) represents the coordinates of the electronic fence boundary point.

[0099] Determine whether the vehicle's final position P(x',y') is within a compliant parking area:

[0100]

[0101] If d' min >δ R The system identifies vehicles as illegally parked and sends an alarm message to the parking management module.

[0102] The beneficial effects of this invention are as follows: This invention adopts a GPS signal quality detection mechanism, calculates the variance of GPS data through multiple samplings, evaluates the positioning accuracy, and directly uses GPS positioning if the signal is stable; otherwise, it calls UWB for position correction and uses a weighted fusion algorithm to calculate the final position information, thereby effectively solving the problem of GPS signal drift in environments such as overpasses and interchanges, and improving the stability of vehicle positioning.

[0103] This invention employs a polygon enclosing algorithm and Euclidean distance calculation for parking area matching to ensure compliant vehicle parking positions. By setting a tolerance range for parking areas, it effectively filters out misjudgments that may be caused by positioning errors, avoiding unnecessary penalties due to GPS or UWB errors. For directional parking, ultra-wide-angle cameras are installed in the parking area, combined with an edge computing unit to perform visual recognition of parked vehicles. The minimum bounding rectangle method is used to calculate the vehicle's principal orientation angle. Simultaneously, combined with IMU sensors and an electronic compass, the orientation information and magnetic azimuth angle are calculated to accurately determine the vehicle's orientation, ensuring that vehicles are parked in the preset direction. If the deviation exceeds a threshold, an adjustment prompt is sent to the user, effectively improving the standardization of directional parking.

[0104] This invention performs a second check on the parking status of a vehicle after parking, acquires an image of the parking area again, and calculates the vehicle's outline to prevent it from being moved by external force after parking, thus avoiding violations. At the parking management level, it further verifies whether the vehicle's position exceeds the parking range through an electronic fence database, and combines secondary visual detection to improve the accuracy of parking monitoring, and triggers alarms when vehicles are parked illegally, thereby improving the level of intelligence in parking management. Attached Figure Description

[0105] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0106] Figure 1 This is a schematic diagram of the framework of the shared electric bicycle fixed-point and directional parking system of the present invention. Detailed Implementation

[0107] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0108] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0109] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0110] Example 1, referring to Figure 1 This embodiment provides a designated parking system for shared electric bicycles, including:

[0111] The vehicle positioning module, including an ultra-wideband (UWB) positioning unit and a GPS signal quality detection unit, is used to obtain the current location of the shared electric bicycle.

[0112] In the vehicle positioning module, the quality of the GPS signal is detected. When the GPS signal meets the preset stability requirements, GPS data is used for positioning. If the GPS signal quality is lower than the preset threshold, the ultra-wideband (UWB) positioning unit is called for position correction.

[0113] The GPS signal quality detection unit is used to determine whether the GPS signal meets the preset stability requirements, and when the GPS signal quality is lower than the threshold, it calls the ultra-wideband UWB positioning unit to perform position correction.

[0114] The parking area matching module is used to determine whether a vehicle is located within a legal parking area;

[0115] The parking area matching module, based on the vehicle location information obtained by the vehicle positioning module, calls the parking area database, compares the current vehicle location with the boundary data of the preset parking area, and determines whether the vehicle is within the compliant parking area.

[0116] The ultra-wide-angle camera in the visual recognition module is set on a fixed bracket or light pole in the parking area, covering the entire parking area, and the image data is processed locally through the edge computing unit;

[0117] The visual recognition module includes an ultra-wide-angle camera and an edge computing unit, which is used to perform visual recognition of parked vehicles in the parking area based on the ultra-wide-angle camera and the edge computing unit to determine whether the vehicle is parked in a preset direction, and to verify the direction data recognized by the user terminal.

[0118] The vehicle orientation detection module, including an inertial measurement unit (IMU) and an electronic compass, is used to acquire vehicle parking orientation data and compare it with the preset orientation standard of the parking area.

[0119] The IMU sensor and electronic compass in the vehicle orientation detection module are integrated into the shared electric bicycle to obtain the vehicle's attitude information and parking angle in real time.

[0120] In the vehicle orientation detection module, if the vehicle position is compliant but the direction deviates from the preset direction, a prompt message to adjust the direction is sent to the user, and the sending method can be the user terminal or the vehicle voice system.

[0121] The vehicle parking verification module performs a secondary verification of the vehicle's parking posture based on the visual recognition module after the user has finished parking. It also uses the edge computing unit to automatically determine the vehicle's parking position and angle, and calls up the electronic fence data to confirm again whether the vehicle's position exceeds the compliant parking area.

[0122] The vehicle parking verification module includes an AI vision-based parking area monitoring unit, which is used to identify whether parked vehicles occupy sidewalks, blind paths, or non-parking areas, and send alarm information to the parking management module when illegal parking occurs;

[0123] The parking management module is used to unlock the settlement function after confirming that the vehicle's location and parking direction are compliant, allowing users to complete the vehicle locking operation and synchronizing the vehicle's parking location, parking direction, and parking time data to the cloud database.

[0124] This embodiment also provides a method for fixed-point and directional parking of the aforementioned shared electric bicycles, including:

[0125] Step S1: After the user rides to the target parking area, the vehicle positioning module obtains the current vehicle location information and detects the GPS signal quality.

[0126] If the GPS signal quality meets the preset stability requirements, then GPS data will be used for vehicle positioning.

[0127] If the GPS signal quality is lower than a preset threshold, the ultra-wideband (UWB) positioning unit will be invoked to correct the vehicle's position.

[0128] The steps for obtaining the current vehicle location information from the vehicle positioning module and simultaneously detecting the GPS signal quality are as follows:

[0129] The vehicle positioning module obtains the current vehicle's GPS data through the GPS signal quality detection unit, which is represented as:

[0130] P GPS =(x GPS ,y GPS ,z GPS ,t GPS SNR GPS ),

[0131] Among them, P GPS This represents the raw location information collected by GPS, x GPS This represents the coordinates of the vehicle's position provided by GPS along the X-axis, y-axis. GPS This represents the coordinates of the vehicle's position along the Y-axis provided by GPS, z. GPS t represents the coordinate value of the vehicle's position in the Z-axis direction provided by GPS. GPS SNR represents the timestamp of a GPS signal. GPS Indicates the signal-to-noise ratio of GPS signals.

[0132] Multiple GPS data collections were performed to calculate the position deviation variance. The calculation formula is:

[0133]

[0134] in, The variance of the GPS location data is represented by N, where N represents the number of GPS data points sampled. The deviation of the j-th GPS data point from the mean is represented by the following formula:

[0135]

[0136] in, These represent the coordinates of the j-th GPS sampling point. These represent the mean coordinates of all GPS sampling points. The average offset distance of the GPS sampling points is expressed by the following formula:

[0137]

[0138] Determine whether the GPS signal quality meets the preset stability requirements, and set the GPS signal quality stability threshold τ. GPS If the following conditions are met, the GPS signal quality is considered stable, and GPS data is used for positioning; otherwise, UWB is used for position correction:

[0139]

[0140] Where, τ GPS SNR is the variance threshold for GPS signal quality. min The minimum acceptable signal-to-noise ratio threshold for GPS signals.

[0141] If the GPS signal quality is below the threshold, UWB will be used for positioning. The UWB location calculation is as follows:

[0142] P UWB =(x UWB ,y UWB ,z UWB ,t UWB ),

[0143] Among them, P UWB Indicates the location information collected by UWB, x UWB ,y UWB ,z UWB Represent the vehicle position coordinates calculated by UWB, t UWB Indicates the UWB collection timestamp.

[0144] The final vehicle position is calculated using a weighted fusion method, and the fusion formula is as follows:

[0145] P = w GPS P GPS +w UWB P UWB ,

[0146] Where P represents the final fused vehicle location information, and w GPS ,w UWB The weights for GPS and UWB are calculated as follows:

[0147]

[0148] Among them, SNR UWBThe signal-to-noise ratio of the UWB signal;

[0149] Specifically, the system collects raw location information via GPS and calculates the location variance based on multiple samplings to assess the stability of the GPS data. If the GPS signal quality meets the preset requirements, the GPS data is used directly for vehicle positioning. If the GPS signal quality is below a set threshold, the ultra-wideband (UWB) positioning unit is called for position correction. The final vehicle position is calculated using a weighted fusion method, which can still provide high-precision location information even when the GPS signal is unstable.

[0150] Step S2: Based on the vehicle location information determined in step S1, call the parking area database, compare the current vehicle location with the preset parking area boundary data, and determine whether the current vehicle is within the compliant parking area.

[0151] The steps to determine whether the current vehicle is within the compliant parking area by comparing its current location with the preset parking area boundary data are as follows:

[0152] Access the parking area database to obtain the boundary data of compliant parking areas. Assume the parking area database provides the boundary information of compliant parking areas.

[0153] R = {(x k ,y k |k=1,2,…,M},

[0154] Where R is the set of boundary points of the compliant parking area, M represents the number of boundary points, (x k ,y k () represents the coordinates of the boundary point.

[0155] An algorithm is used to determine whether a vehicle's position P(x,y) is within a compliant parking area by identifying points inside a polygon. The number of times the boundary is crossed is calculated using the following formula:

[0156]

[0157] Where C represents the number of times the vehicle crosses the boundary. If C is odd, the vehicle is within the compliant parking area; otherwise, the vehicle is outside the compliant area.

[0158] The minimum Euclidean distance between the vehicle's current position and the boundary point of the parking area is calculated using the following formula:

[0159]

[0160] Where, d min This indicates the shortest distance from the vehicle's location to the boundary of the compliant parking area.

[0161] Let the tolerance range of the parking area be δ. R If the following conditions are met:

[0162] d min ≤δ R If the vehicle is parked within the permitted parking area, it is considered to be in a compliant parking area; otherwise, the vehicle is parked improperly.

[0163] Specifically, this step first calls the parking area database to obtain the boundary data of compliant parking areas; then, based on the vehicle location information determined in step S1, a polygon enclosing algorithm is used to determine whether the vehicle is located within the compliant parking area; if the vehicle is located inside the area, the parking location is considered compliant; if the vehicle exceeds the boundary of the parking area, the minimum distance between the vehicle and the boundary is further calculated and compared with the tolerance range of the parking area to determine whether the vehicle complies with the parking regulations.

[0164] Effectively filters out positioning errors that may exist in GPS or UWB, improving the ability to detect illegal parking behavior;

[0165] Step S3: Within the parking area, visual recognition of parked vehicles is performed based on an ultra-wide-angle camera and an edge computing unit to determine whether the vehicles are parked in a preset direction.

[0166] Vehicle parking direction data is acquired using an onboard inertial measurement unit (IMU) and an electronic compass, and compared with the preset direction standard of the parking area.

[0167] If the vehicle is in the correct position but deviates from the preset direction, a prompt message will be sent to the user to adjust the direction.

[0168] The steps for visually recognizing parked vehicles and determining whether they are parked in a preset direction using an ultra-wide-angle camera and edge computing unit are as follows:

[0169] The ultra-wide-angle camera in the visual recognition module acquires image data of the parking area, represented as follows:

[0170] I = {I f |f=1,2,…,F},

[0171] Where I is the image dataset of the parking area, I f This represents the f-th frame of the image, where F is the total number of image frames.

[0172] The edge computing unit is used to extract the contour of the image data, specifically the contour region of the parked vehicle. The extraction formula is as follows:

[0173] C V ={(x c ,y c |c=1,2,…,C},

[0174] Among them, C VThe set of pixels representing the vehicle outline, where C is the total number of outline pixels, (x c ,y c () represents the pixel coordinates of the c-th contour point.

[0175] The vehicle's principal steering angle is calculated using the minimum bounding rectangle method. The formula is as follows:

[0176]

[0177] Where, θ V The vehicle heading angle for visual recognition, x max ,x min The coordinates of the rightmost and leftmost points of the contour boundary, y max ,y min The coordinates of the uppermost and lowermost points of the contour boundary.

[0178] The vehicle orientation detection module acquires vehicle attitude information P through an IMU sensor. IMU :

[0179] P IMU =(α IMU ,β IMU ,γ IMU ,t IMU ),

[0180] Among them, P IMU α represents the attitude information acquired by the IMU. IMU ,β IMU ,γ IMU These represent the vehicle's roll angle, pitch angle, and yaw angle, respectively. IMU Indicates the timestamp of IMU data.

[0181] The electronic compass acquires magnetic field data and calculates the vehicle's magnetic azimuth angle using the following formula:

[0182]

[0183] Where, θ M B represents the vehicle heading angle calculated by the electronic compass. x B y These represent the magnetic induction intensity of the magnetic field sensor along the X and Y axes, respectively.

[0184] Compare whether the vehicle's steering angle conforms to the preset direction. Let θ be the preset direction standard for the parking area. std Determine the vehicle's directional deviation:

[0185] Δθ=min(|θ V -θ std |,|θ M -θ std |),

[0186] Where Δθ represents the angle of deviation between the vehicle's direction and the preset direction.

[0187] If Δθ exceeds the threshold θ max The user terminal and the vehicle's voice module send adjustment prompts;

[0188] Specifically, this step uses an ultra-wide-angle camera to acquire image data of the parking area, and uses an edge computing unit to extract the vehicle outline. The vehicle's heading angle is calculated using the minimum bounding rectangle method, and the vehicle's attitude and magnetic azimuth are calculated by combining data from the onboard inertial measurement unit (IMU) and electronic compass. The acquired vehicle direction data is compared with the preset direction standard of the parking area to determine whether the vehicle is parked in the prescribed direction.

[0189] If the deviation exceeds the preset threshold, a prompt message to adjust the direction will be sent to the user terminal and the vehicle voice module.

[0190] Step S4: After the user finishes parking, the vehicle's parking posture is checked a second time based on the visual recognition module, and the parking position and angle of the vehicle are automatically determined by the edge computing unit.

[0191] In step S3, based on the vehicle location information determined in step S1, and by calling the electronic fence data, it is confirmed again whether the vehicle location exceeds the compliant parking area.

[0192] The steps involve a secondary verification of the vehicle's parking posture using a visual recognition module, followed by automatic determination of the vehicle's parking position and angle using an edge computing unit.

[0193] After the user completes parking, the visual recognition module re-captures the parking area image I':

[0194] I'={I' f |f=1,2,…,F'},

[0195] Where I' represents the image dataset during the parking review phase, I' f This represents the f-th frame of the verification image, where F' is the total number of verification image frames.

[0196] The vehicle outline is identified again, and the vehicle outline for the review stage is calculated using the following formula:

[0197] C' V ={(x' c ,y' c |c=1,2,…,C'},

[0198] Among them, C' VLet C' be the set of pixels representing the vehicle outline during the verification phase, and let C' be the total number of outline pixels during the verification phase.

[0199] The formula for calculating the vehicle steering angle during the second verification stage is as follows:

[0200]

[0201] Where, θ' V x' represents the vehicle orientation angle for visual recognition during the second verification stage. max ,x' min During the review phase, the coordinates of the rightmost and leftmost ends of the vehicle outline boundary, y' max ,y' min The coordinates of the top and bottom edges of the vehicle's outline boundary during the verification phase.

[0202] Accessing the electronic fence data confirms the location's compliance; the electronic fence database provides boundary information for the parking area. E :

[0203] R E ={(x m ,y m )|m=1,2,…,M'},

[0204] Among them, R E M' is the set of boundary points for the parking area defined by the electronic fence, where M' is the number of electronic fence boundary points. m ,y m ( ) represents the coordinates of the electronic fence boundary point.

[0205] Determine whether the vehicle's final position P(x',y') is within a compliant parking area:

[0206]

[0207] If d' min >δ R The system identifies vehicles as illegally parked and sends an alarm message to the parking management module.

[0208] Specifically, the vehicle's parking posture is verified a second time based on the visual recognition module, and the parking position and angle are automatically determined by the edge computing unit to improve the accuracy of parking monitoring.

[0209] By comparing the initial detection results with the image data from the review stage, the system ensures that vehicles remain compliantly parked when finally locked. Simultaneously, it accesses the electronic fence database to obtain parking area boundary data and calculates the minimum distance between the vehicle's location and the boundary to further confirm whether the vehicle has exceeded the compliant parking area. If a vehicle is illegally parked, an alarm is sent to the parking management module, thus improving the intelligence level of parking management.

[0210] In step S5, after the parking management module confirms that the vehicle location and parking direction are compliant, it unlocks the settlement function, allowing the user to complete the vehicle locking operation, and synchronizes data such as vehicle parking location, parking direction, and parking time to the cloud database.

[0211] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A shared electric motorcycle fixed-point directional parking system, characterized by: include, The vehicle positioning module includes an ultra-wideband (UWB) positioning unit and a GPS signal quality detection unit, which are used to obtain the current location of the shared electric motorcycle; The parking area matching module is used to determine whether the vehicle is within the legal parking range; A visual recognition module, including an ultra-wide-angle camera and an edge computing unit, is used to perform visual recognition of parked vehicles in a parking area based on the ultra-wide-angle camera and the edge computing unit to determine whether the vehicle is parked in a preset direction, and to verify the direction data recognized by the user end; The vehicle orientation detection module includes an inertial measurement unit (IMU) and an electronic compass, which is used to obtain the vehicle parking direction data and compare it with the preset direction standards of the parking area; The vehicle parking review module conducts a secondary review of the vehicle's parking posture based on the visual recognition module after the user completes parking. It automatically determines the vehicle's parking position and parking angle in combination with the edge computing unit, and calls the electronic fence data to confirm again whether the vehicle's position exceeds the scope of the compliant parking area. The parking management module is used to unlock the settlement function after confirming that the vehicle position and parking direction are compliant, allowing the user to complete the vehicle locking operation and synchronize the vehicle parking position, parking direction and parking time data to the cloud database.

2. A shared electric motorcycle fixed-point directional parking system as claimed in claim 1, characterized in that: In the vehicle positioning module, the quality of the global satellite navigation system GPS signal is detected. When the GPS signal meets the preset stability requirements, GPS data is used for positioning. If the GPS signal quality is lower than the preset threshold, the ultra-wideband UWB positioning unit is called for position correction. The GPS signal quality detection unit is used to determine whether the GPS signal meets the preset stability requirement, and when the GPS signal quality is lower than a threshold, call the ultra-wideband (UWB) positioning unit to perform position correction.

3. A shared electric motorcycle fixed-point directional parking system as claimed in claim 2, characterized in that: The parking area matching module calls the parking area database based on the vehicle position information obtained by the vehicle positioning module, compares the current vehicle position with the boundary data of the preset parking area, and determines whether the vehicle is within the compliant parking range; The ultra-wide-angle camera in the visual recognition module is set on a fixed bracket or lamp pole in the parking area, covering the entire parking area, and the image data is locally processed through the edge computing unit.

4. A shared electric motorcycle fixed-point directional parking system as claimed in claim 3, characterized in that: The IMU sensor and electronic compass in the vehicle orientation detection module are integrated into the shared electric motorcycle to obtain the vehicle's posture information and parking angle in real time; In the vehicle orientation detection module, if the vehicle position is in compliance but the direction deviates from the preset direction, a prompt message for adjusting the direction is sent to the user.

5. A shared electric motorcycle fixed-point directional parking system as claimed in claim 4, characterized in that: The vehicle parking review module includes an AI vision-based parking area monitoring unit, which is used to identify whether the parked vehicle occupies the sidewalk, blind path or non-parking area, and sends an alarm message to the parking management module when the parking is illegal.

6. A shared electric motorcycle fixed-point directional parking system as claimed in claim 5, characterized in that: The parking method of this system is: Step S1: After the user rides to the target parking area, the vehicle positioning module obtains the current vehicle location information and detects the GPS signal quality; If the GPS signal quality meets the preset stability requirements, the GPS data is used to locate the vehicle; If the GPS signal quality is lower than the preset threshold, the ultra-wideband (UWB) positioning unit is called to correct the vehicle position; Step S2, based on the vehicle position information determined in step S1, calling the parking area database, comparing the current vehicle position with the preset parking area boundary data, and determining whether the current vehicle is located within the compliant parking range; Step S3: In the parking area, visual recognition of parked vehicles is performed based on the ultra-wide-angle camera and the edge computing unit to determine whether the vehicle is parked in a preset direction; The vehicle's parking direction data is obtained using the vehicle's inertial measurement unit (IMU) and electronic compass, and compared with the preset direction standards of the parking area; If the vehicle's position is in compliance but the direction deviates from the preset direction, a prompt message is sent to the user to adjust the direction; Step S4: After the user completes parking, the parking posture of the vehicle is rechecked based on the visual recognition module, and the parking position and parking angle of the vehicle are automatically determined in combination with the edge computing unit; In step S3, based on the vehicle location information determined in step S1 and calling the electronic fence data, it is confirmed again whether the vehicle location exceeds the scope of the compliant parking area; Step S5: After the parking management module confirms that the vehicle position and parking direction are in compliance, it unlocks the settlement function, allowing the user to complete the vehicle locking operation, and synchronizes the vehicle parking position, parking direction, parking time and other data to the cloud database.

7. A shared electric motorcycle fixed-point directional parking system as claimed in claim 6, characterized in that: The step of obtaining the current vehicle location information by the vehicle positioning module and detecting the signal quality of the GPS is as follows: The vehicle positioning module obtains the GPS data of the current vehicle through the GPS signal quality detection unit, which is expressed as: P GPS =(x GPS ,y GPS ,z GPS ,t GPS ,SNR GPS ), Among them, P GPS Indicates the original location information collected by GPS, x GPS Indicates the coordinate value of the vehicle position in the X-axis direction provided by GPS, y GPS Indicates the coordinate value of the vehicle position in the Y-axis direction provided by GPS, z GPS Indicates the coordinate value of the vehicle position in the Z-axis direction provided by GPS, t GPS Indicates the timestamp of the GPS signal, SNR GPS represents the signal-to-noise ratio of the GPS signal, Collect GPS data multiple times and calculate the position deviation variance The calculation formula is: in, represents the variance of GPS position data, N represents the number of sampled GPS data points, It represents the deviation of the jth GPS data point from the mean, and the calculation formula is: in, Respectively represent the coordinate values ​​of the jth GPS sampling point, Respectively represent the mean coordinates of all GPS sampling points, Indicates the average offset distance of the GPS sampling points, and the calculation formula is: Determine whether the GPS signal quality meets the preset stability requirements and set the stability threshold τ of the GPS signal quality GPS If the following conditions are met, the GPS signal quality is considered stable and GPS data is used for positioning. Otherwise, UWB is called for position correction: SNR GPS >SNR min , Among them, τ GPS is the variance threshold of GPS signal quality, SNR min The minimum acceptable signal-to-noise ratio threshold of the GPS signal, If the GPS signal quality is lower than the threshold, UWB is used for positioning instead. The UWB calculated position is: P UWB =(x UWB ,y UWB ,z UWB ,t UWB ), Among them, P UWB Indicates the location information collected by UWB, x UWB ,y UWB ,z UWB They represent the vehicle position coordinates calculated by UWB, t UWB Indicates the UWB acquisition timestamp, The weighted fusion method is used to calculate the final vehicle position. The fusion formula is: P=w GPS P GPS +w UWB P UWB , Among them, P is the final fused vehicle position information, w GPS ,w UWB is the weight of GPS and UWB, calculated as: Among them, SNR UWB is the signal-to-noise ratio of the UWB signal.

8. A shared electric motorcycle fixed-point directional parking system as claimed in claim 7, characterized in that: The step of comparing the current vehicle position with the preset parking area boundary data to determine whether the current vehicle is within the compliant parking range is: Call the parking area database to obtain the boundary data of the compliant parking area. Assume that the parking area database provides the boundary information of the compliant parking area: R={(x k ,y k )|k=1,2,…,M}, Where R is the set of boundary points of the compliant parking area, M is the number of boundary points, (x k ,y k ) are the boundary point coordinates, The point-in-polygon judgment algorithm is used to determine whether the vehicle position P(x, y) is located in the legal parking area and the number of times the boundary is crossed is calculated. The calculation formula is: Where C is the number of times the boundary is crossed. If C is an odd number, the vehicle is within the compliance parking range. Otherwise, the vehicle is beyond the compliance range. Calculate the minimum Euclidean distance between the vehicle's current position and the parking area boundary point. The calculation formula is: Among them, d min Indicates the shortest distance from the vehicle position to the boundary of the legal parking area. Assume the tolerance range of the parking area is δ R , if it satisfies: d min ≤δ R , the vehicle is considered to be in a compliant parking area, otherwise the vehicle is parked in an ineligible manner.

9. A shared electric motorcycle fixed-point directional parking system as claimed in claim 8, characterized in that: The step of visually identifying the parked vehicle based on the ultra-wide-angle camera and the edge computing unit to determine whether the vehicle is parked in a preset direction is as follows: The ultra-wide-angle camera of the visual recognition module obtains image data of the parking area, which is expressed as: I={I f |f=1,2,…,F}, Among them, I is the image dataset of the parking area, I f represents the f-th frame image, F is the total number of image frames, Use the edge computing unit to perform contour extraction on the image data and extract the contour area of ​​the parked vehicle. The extraction formula is: C V ={(x c ,y c )|c=1,2,…,C}, Among them, C V The pixel set representing the vehicle contour, C is the total number of contour pixels, (x c ,y c ) represents the pixel coordinates of the cth contour point, The minimum circumscribed rectangle method is used to calculate the main direction angle of the vehicle. The calculation formula is: Among them, θ V represents the direction angle of the vehicle identified by vision, x max ,x min The rightmost and leftmost coordinates of the contour boundary, y max ,y min The top and bottom coordinates of the contour boundary, The vehicle orientation detection module obtains the vehicle posture information P through the IMU sensor IMU : P IMU =(a IMU ,b IMU ,c IMU ,t IMU , Among them, P IMU express I The posture information collected by MU, α IMu ,β IMU ,γ IMU Respectively represent the roll angle, pitch angle and yaw angle of the vehicle, t IMU Indicates the timestamp of the IMU data, The electronic compass obtains magnetic field data and calculates the vehicle's magnetic azimuth. The calculation formula is: Among them, θ M Indicates the vehicle direction angle calculated by the electronic compass, B x ,B y Respectively represent the magnetic induction intensity of the magnetic field sensor on the X-axis and Y-axis, Compare the vehicle direction angle to see if it meets the preset direction. Suppose the preset direction standard of the parking area is θ std , determine the vehicle direction deviation: Δθ=min(|θ V -θ std| ,|θ M -θ std|) , Among them, Δθ represents the deviation angle between the vehicle direction and the preset direction, If Δθ exceeds the threshold θ max The user terminal and the vehicle-mounted voice module send adjustment prompt information.

10. A shared electric motorcycle fixed-point directional parking system as claimed in claim 9, characterized in that: The steps of performing a secondary review of the parking posture of the vehicle based on the visual recognition module and automatically determining the parking position and parking angle of the vehicle in combination with the edge computing unit are as follows: After the user completes parking, the visual recognition module re-collects the parking area image I': I'={I' f |f=1,2,…,F'}, Among them, I' represents the image dataset of the parking review stage, I' f represents the f-th frame of the review image, F' is the total number of review image frames, Identify the vehicle outline again and calculate the vehicle outline in the review stage. The calculation formula is: C' V ={(x' c ,y' c )|c=1,2,…,C'}, Among them, C' V is the pixel set of the vehicle contour in the review stage, C' is the total number of contour pixels in the review stage, Calculate the vehicle direction angle in the second review stage using the following formula: Among them, θ' V represents the vehicle direction angle of the visual recognition in the second review stage, x' max ,x' min The rightmost and leftmost coordinates of the vehicle contour boundary during the review phase, y' max ,y' min The top and bottom coordinates of the vehicle contour boundary during the review phase, Call the electronic fence data to confirm whether the location is compliant. The electronic fence database provides the boundary information of the parking area. E : R E ={(x m ,y m )|m=1,2,…,M'}, Among them, R E is the parking area boundary point set defined by the electronic fence, M' is the number of electronic fence boundary points, (x m ,y m ) are the coordinates of the boundary points of the electronic fence, Determine whether the final position P(x',y') of the vehicle is within the legal parking area: If d' min >δ R Identify vehicles that are parked illegally and send warning information to the parking management module.

Images