Vehicle-cloud integrated vehicle track data adaptive processing method

Through the vehicle-cloud-integrated vehicle trajectory data adaptive processing method, the problems of missing and insufficient real-time in data acquisition and processing are solved. Through redundant merging, null value filling and outlier point correction technologies, the real-time vehicle trajectory data is improved. and accuracy, especially data quality under special operating conditions, supports the efficient operation of intelligent transportation systems.

CN119089118AActive Publication Date: 2024-12-06SUN YAT SEN UNIV

Patent Information

Application Number
CN202411261470.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-10
Publication Date
2024-12-06
Estimated Expiration
2044-09-10

AI Technical Summary

Technical Problem

The prior art has problems such as data loss, insufficient real-time and neglect of outliers in the acquisition and processing of vehicle trajectory data. Especially in special operating conditions, data analysis is not in-depth enough, which affects the rationality of data and the accuracy of trajectory prediction.

Method used

Adaptive processing methods for vehicle trajectory data integrated into the vehicle and cloud include data acquisition, redundant merging, empty value filling, outlier point correction and other steps. Data is collected through multiple devices, set the time window to merge redundant data, and use it Adapt window technology fills the null values, combines sample entropy and local outliers to identify outliers, improving the real-time and accuracy of the data.

Benefits of technology

Effectively avoid data loss, improve the real-time and rationality of vehicle trajectory data, enhance data quality under special operating conditions, reduce the burden of redundant data processing, improve the accuracy of trajectory prediction and the timeliness of autonomous driving decisions .

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119089118A_ABST
    Figure CN119089118A_ABST
Patent Text Reader

Abstract

The invention provides a vehicle-cloud integrated vehicle trajectory data adaptive processing method, which comprises the following steps: S1, data acquisition: at a vehicle end, using various data acquisition devices to acquire vehicle trajectory data, the acquired vehicle trajectory data comprising a space coordinate and a timestamp of a vehicle, the collected vehicle track data is transmitted to a computing device of the vehicle through a data bus in the vehicle or a wireless communication mode; s2, data preprocessing including sub-steps of redundancy merging, null value filling and the like; and S3, data association, including sub-steps of spatial association, data slicing, outlier correction and the like. According to the invention, the vehicle track data is obtained by using multiple means, the missing of the vehicle track data is effectively avoided, and the real-time performance of the vehicle track data is improved. By adopting the technical means of outlier correction, redundancy merging and the like, the rationality of vehicle trajectory data, the accuracy of trajectory prediction, the processing efficiency of the vehicle trajectory data and the real-time performance of feedback are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of intelligent transportation, and specifically relates to a vehicle-cloud integrated vehicle trajectory data adaptive processing method. Background Art

[0002] The invention patent application with application number 202311582372.5 discloses a multimodal analysis method for autonomous driving vehicle trajectory data. The basic technical content of the method includes: obtaining the instantaneous linear quantity and instantaneous angular quantity of each vehicle, wherein the instantaneous linear quantity includes the instantaneous coordinate position, the instantaneous driving speed, and the instantaneous driving acceleration; the instantaneous angular quantity includes the instantaneous azimuth, the instantaneous angular velocity, and the instantaneous angular acceleration; establishing a Gaussian distribution function of the instantaneous linear quantity and the instantaneous angular quantity, and determining the motion mode of each vehicle based on the relationship between the data mean and variance in the Gaussian distribution function of the instantaneous linear quantity and the instantaneous angular quantity.

[0003] The invention patent application with application number 202310331776.0 discloses a vehicle abnormal data screening method for trajectory data, which includes the following main steps: (1) obtaining a highway vehicle trajectory data set and preprocessing the vehicle trajectory data set; (2) judging the preprocessed vehicle trajectory data set, if the sampling time interval is lower than a preset threshold, then entering step (3), otherwise entering step (4); (3) resampling the vehicle trajectory data set; (4) dividing the vehicle trajectory data set into sub-data sets according to time characteristics and spatial characteristics; (5) analyzing the data characteristics of the vehicles in each sub-data set, and marking the abnormal indicator data as abnormal data according to the analysis results of the data characteristics.

[0004] The above inventions each have their own advantages. Specifically, the technical solution recorded in the invention patent application with application number 202311582372.5 deeply considers the characteristics and features of the movement of autonomous vehicles, and constructs a multimodal analysis and verification method for autonomous driving trajectory data, which can classify the vehicle's movement behavior under different conditions, and provide accurate and reliable data support for key technologies such as path planning and control decision-making of autonomous driving systems. The technical solution recorded in the invention patent application with application number 202310331776.0 divides the trajectory data into sub-data sets based on time and space characteristics, and performs feature analysis on the three indicators of vehicle speed, acceleration, and headway in the sub-data sets, reducing the interference of time and space characteristics, thereby effectively screening out abnormal data that occurs during vehicle driving.

[0005] However, the above invention also has some technical defects, and its common technical defects are mainly manifested in the following two aspects:

[0006] First, the means of acquiring vehicle trajectory data are relatively limited, resulting in partial data missing or poor real-time performance, and data missing will affect the integrity and accuracy of the vehicle trajectory. In particular, the above invention ignores the outliers that often exist in vehicle trajectory data, thereby affecting the rationality of the data and the accuracy of trajectory prediction. In addition, the above invention does not conduct in-depth data analysis for special working conditions (such as vehicle turning) and lacks methods that can be adaptively processed, which seriously affects the data quality under special working conditions.

[0007] Second, the above invention will generate a large amount of redundant data during the data processing process. A large amount of redundant data will increase the burden of transmission and processing, resulting in insufficient real-time data processing and feedback, seriously affecting the timeliness of autonomous driving decisions. Summary of the invention

[0008] The purpose of the present invention is to overcome the first and second technical deficiencies of the above-mentioned prior art.

[0009] In order to achieve the above object, the present invention adopts the following technical solutions:

[0010] A vehicle-cloud integrated vehicle trajectory data adaptive processing method comprises the following steps:

[0011] S1, data collection: at the vehicle end, a data collection device is used to collect vehicle trajectory data, the collected vehicle trajectory data includes the spatial coordinates and timestamp of the vehicle, and the collected vehicle trajectory data is transmitted to the vehicle's computing device via the vehicle's internal data bus or wireless communication;

[0012] S2, data preprocessing: the vehicle trajectory data is preliminarily processed by the vehicle's computing device, including the following sub-steps:

[0013] S201, redundant merging: setting a time window for redundant data, and merging vehicle trajectory points whose time difference is less than the time window;

[0014] S202, null value filling: using adaptive window technology to dynamically adjust the sampling window according to the vehicle trajectory data density;

[0015] S203, data upload: uploading the pre-processed vehicle trajectory data to the cloud via a wireless network;

[0016] S3, data association; includes the following sub-steps:

[0017] S301, spatial association; using geographic spatial index to realize spatial data retrieval and management, spatially associating the redundant merged and null-filled vehicle trajectory points with the road network data, traffic element information and regional POI information;

[0018] S302, data slicing: slicing the geographic spatial data according to a preset grid size to achieve grid management of spatial data;

[0019] S303, outlier correction: using an adaptive outlier screening algorithm based on sample entropy and local outlier factor to identify outliers in the vehicle trajectory;

[0020] S304, generating vehicle trajectory information: merging the corrected vehicle trajectory points with the normal point set to generate vehicle trajectory information.

[0021] On the basis of the above technical solutions, the present invention may add the following technical means to better or more specifically solve the technical problems to be solved by the present invention:

[0022] The data acquisition equipment includes video monitoring, ground sensing coils, wireless radio frequency identification, automotive electronic identification, and on-board diagnostic system equipment and global positioning system equipment installed in smart vehicles.

[0023] Further, when executing S201, the vehicle trajectory data is set to: {(x i ,y i ,t i )},x i and i represents the spatial coordinates of the vehicle, t i Represents the timestamp, sets a time window Δt for redundant data, merges the vehicle trajectory data points with a time difference less than Δt, and the merged vehicle trajectory point position is the weighted average of all data points within the time window Δt:

[0024]

[0025] In the formula, w i is the weight, w i The value is set to 1 or according to the reliability of the data point.

[0026] Furthermore, when S202 is executed under special working conditions (such as vehicle turning), encrypted data points are used to perform complexity calculation on the vehicle trajectory data based on the concept of entropy. For the two-dimensional trajectory coordinates, vector difference is used to perform sample entropy analysis to describe the complexity of instantaneous changes in the vehicle trajectory and capture nonlinear characteristics in the signal. By judging the degree of change of sample entropy within a certain range, a threshold is set to adaptively and dynamically adjust the window size and fill in empty values.

[0027] Further, when S202 is executed, the following steps are included:

[0028] First, a sampling window with an initial size of L is set, which slides across the vehicle trajectory sequence and calculates the sample entropy of the vehicle trajectory vector in each window segment;

[0029] Secondly, a threshold is set and the sample entropy of adjacent segments is subtracted in pairs; the segments whose subtraction results exceed the threshold are regarded as segments with drastic changes in the curvature of the vehicle trajectory;

[0030] Next, obtain the maximum length Lmax that meets the threshold judgment condition, and calculate the sample entropy mean of the vehicle trajectory corresponding to 1-Lmax;

[0031] Next, sort the sample entropy means from small to large, draw a scatter plot and connect the minimum and maximum values ​​to form a straight line;

[0032] Finally, the distance l from each point to the straight line is calculated in turn, and the horizontal coordinate corresponding to the maximum distance lmax is the optimal window length L.

[0033] Furthermore, for a given vehicle trajectory sequence B = {b1, b2, K, b n}, the sample entropy is calculated as follows:

[0034]

[0035] In the formula, the integer 0≤m≤N represents the length of the sequence to be compared, the real number r>0 represents the tolerance for accepting a match, and C m (r) represents the empirical probability of two sequences matching m points, which is obtained by the following formula:

[0036]

[0037] In the formula, Obtained by the following formula:

[0038]

[0039] In the formula, Represents any vector x m (j) in x m (i) The probability within the range r, NumT(x m (i),x m (j)) represents x m (i) and x m (j) is less than the number of times r, x m (i) and x m (j) is the m-dimensional vector reconstructed by B.

[0040] Furthermore, linear interpolation is used to fill in the empty values. Let the two known vehicle trajectory points be (x i-1 ,y i-1 ,t i-1 ) and (x i+1 ,y i+1 ,ti+1 ), the missing point in the middle is (x i ,y i ,t i ),but:

[0041]

[0042] After obtaining L, the time window slides a certain step size and begins to adaptively process the next trajectory sequence.

[0043] Furthermore, when executing S302, the entire geographic space is divided into n×n grids, each grid has a size of Δx×Δy, and the network index to which the vehicle trajectory point (x, y) belongs is:

[0044]

[0045] Furthermore, when executing S303, the degree of outlier of the data point in the local area is adaptively calculated according to the density of neighboring points around the data point, including the following steps:

[0046] First, find the k nearest points of the target point d and calculate their reachability distance RD (reachability distance). The definition of reachability distance RD is as follows:

[0047] RD=max{k dis ,d dis (d,d j )},

[0048] In the formula, k dis is the distance from point d to the kth nearest neighbor; d dis (d,d j ) is point d to d j distance (j∈{1,2,…,k});

[0049] Secondly, traverse the vehicle trajectory and calculate the RD mean reciprocal (LRD) of each point in turn and store it in set R1;

[0050] Next, traverse the trajectory again and calculate the LOF of each point in combination with R1 and store it in the set L1, and we get:

[0051]

[0052] Finally, L1 is judged and screened according to the threshold c to obtain the final target outlier point set.

[0053] Furthermore, the vehicle trajectory point set is set to {(x i ,y i ,t i )}, the outlier is (x out,y out ,t out ), the final target outlier point can be corrected by using the two outliers before and after, and the relationship is as follows:

[0054]

[0055] Furthermore, the cloud analyzes and models the historical trajectory data to generate predicted trajectories.

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] First, on the vehicle side, the present invention uses multiple means to obtain vehicle trajectory data, which not only effectively avoids the loss of vehicle trajectory data, but also improves the real-time performance of vehicle trajectory data.

[0058] Second, in the processing of vehicle trajectory data, great attention is paid to the outliers of vehicle trajectory data. By adopting technical means such as outlier correction, the rationality of vehicle trajectory data and the accuracy of trajectory prediction are improved.

[0059] Third, by adopting the technical means of redundant merging, the efficiency of vehicle trajectory data processing and the real-time feedback are improved.

[0060] Fourth, for special working conditions (such as vehicle turning and other driving conditions), encrypted data points are used, and the complexity of vehicle trajectory data is calculated based on the concept of entropy, which improves the quality of vehicle trajectory data under special working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a flowchart of an embodiment of the present invention. DETAILED DESCRIPTION

[0062] In order to facilitate those skilled in the art to fully understand the technical solution of the present invention, an embodiment of the present invention is described below with reference to the accompanying drawings.

[0063] like Figure 1 As shown, a vehicle trajectory data adaptive processing method for vehicle-cloud integration includes the following steps:

[0064] S1, data collection;

[0065] On the vehicle side (referred to as the vehicle side), a variety of vehicle trajectory data collection devices are used, such as video surveillance cameras, loop detectors, radio frequency identification (RFID), electronic registration identification (ERI), etc., as well as on-board diagnostics (OBD) devices and global positioning system (GPS) devices installed in smart vehicles to collect vehicle trajectory data. The collected vehicle trajectory data includes the spatial coordinates (longitude and latitude) and timestamp of the vehicle. The collected vehicle trajectory data is transmitted to the vehicle's computing device through the vehicle's internal data bus or wireless communication method, and pre-processed by the vehicle's computing device.

[0066] S2, data preprocessing;

[0067] Data preprocessing, that is, the vehicle's computing device preprocesses the received vehicle trajectory data, with the purpose of reducing the amount of data transmission. Vehicle trajectory data preprocessing includes the following sub-steps:

[0068] S201, redundant merging;

[0069] Set the vehicle trajectory data to {(x i ,y i ,t i )}, where x i and i represents the spatial coordinates of the vehicle, t i Represents the timestamp. For redundant data, a time window Δt can be set to merge the vehicle trajectory data points with a time difference less than Δt (the merging of vehicle trajectory data points corresponds to the merging of vehicle trajectory points). The position of the merged vehicle trajectory point can be the weighted average of all vehicle trajectory data points in the window, as shown below:

[0070]

[0071] In the formula, w i is the weight, usually taken as 1 or set according to the reliability of the data point.

[0072] S202, null value filling;

[0073] Adaptive window technology is used to dynamically adjust the sampling window according to the data density to improve the accuracy of null value filling. It is worth noting that under special working conditions (such as vehicle turning), more dense data points are required to obtain more detailed trajectory changes. To this end, this embodiment is based on the concept of entropy, performs complexity calculation on vehicle trajectory data, uses vector difference to analyze sample entropy for two-dimensional trajectory coordinates, describes the complexity of instantaneous changes in trajectory, and better captures nonlinear characteristics in signals. On this basis, by judging the degree of change of sample entropy within a certain range, a threshold is set to adaptively and dynamically adjust the window size and fill in null values, thereby enhancing the local change characteristics of the vehicle trajectory, avoiding information loss, and ensuring the continuity and integrity of the vehicle trajectory.

[0074] In this embodiment, the specific steps of filling in null values ​​are as follows:

[0075] First, a sampling window with an initial size of L is set, which slides across the vehicle trajectory sequence and calculates the sample entropy of the vehicle trajectory vector in each window segment;

[0076] Secondly, a threshold is set and the sample entropy of adjacent segments is subtracted in pairs. The segments whose subtraction results exceed the threshold are regarded as segments with drastic changes in the curvature of the vehicle trajectory.

[0077] Next, obtain the maximum length Lmax that meets the threshold judgment condition, and calculate the sample entropy mean of the vehicle trajectory corresponding to 1-Lmax;

[0078] Next, sort these values ​​from small to large, draw a scatter plot and connect the minimum and maximum values ​​to form a straight line;

[0079] Finally, the distance l from each point to the line is calculated in turn, and the horizontal coordinate corresponding to the maximum distance lmax is the optimal window length L. In order to ensure that the window can cover the entire special working condition trajectory and ensure the integrity of the short-term trajectory characteristics. For a given vehicle trajectory sequence B = {b1, b2, K, b n}, the sample entropy is calculated as follows:

[0080]

[0081] In the formula, the integer 0≤m≤N represents the length of the sequence to be compared, the real number r>0 represents the tolerance for accepting a match, and C m (r) represents the empirical probability of two sequences matching m points, which is obtained by the following formula:

[0082]

[0083] In the formula, It is obtained by the following formula:

[0084]

[0085] In the formula, It can be viewed as any vector x m (j) in x m (i) The probability within the range r, NumT(x m (i),x m (j)) represents x m (i) and x m (j) The number of times a certain distance relationship (such as Euclidean distance) is less than r, x m (i) and x m (j) is the m-dimensional vector reconstructed by B. When processing data, linear interpolation is used to fill in the empty values. Let the two known trajectory points be (x i-1 ,y i-1 ,t i-1 ) and (x i+1 ,y i+1 ,t i+1 ), the missing point in the middle is (x i ,y i ,t i ),but:

[0086]

[0087] After obtaining L, the time window slides a certain step length and begins to adaptively process the next segment of the trajectory sequence. When the entire segment is processed, the optimal window length set L can be obtained. A According to L A By dividing the trajectory, we can obtain complete fragments of change information.

[0088] S203, data upload;

[0089] The vehicle uploads the pre-processed vehicle trajectory data to the cloud via a wireless network (such as 4G / 5G). The data transmission volume can be estimated by the following formula:

[0090]

[0091] Where D t is the total data transmission volume, S t is the size of the ith data packet, and n is the number of data packets

[0092] S3, data association;

[0093] Data association includes the following sub-steps:

[0094] S301, spatial association;

[0095] By using geospatial indexes (such as R-trees) to achieve efficient spatial data retrieval and management, the corrected vehicle trajectory points (i.e., vehicle trajectory points after redundant merging and null value filling) can be quickly spatially associated with road network data, traffic element information, and regional POI information. It can also accurately locate and associate vehicle trajectory points with other elements in the geographic space (such as roads, traffic signs, and points of interest), thus achieving refined management and analysis of vehicle travel paths.

[0096] Assume the trajectory point of the vehicle is (x, y), and the spatial data is {(x j ,y j ,attributes j )}, find the spatial data closest to the vehicle trajectory point:

[0097]

[0098] Through nearest neighbor search or range query, the vehicle trajectory data is associated with a variety of spatial information to provide a comprehensive spatial data view to support complex traffic management and analysis needs. Through geospatial indexing, the vehicle trajectory points are accurately matched with road network data and traffic elements to ensure the spatial accuracy of the trajectory data, which helps to better understand and manage the actual driving routes of vehicles.

[0099] S302, data slicing;

[0100] Slice the geospatial data according to the preset grid size to achieve grid management of spatial data. By dividing the entire space into n×n grids, each grid size is Δx×Δy, large-scale geospatial data is divided into smaller, easier-to-manage units, significantly improving the efficiency of data processing and query. Each grid unit can be processed independently, reducing computational complexity.

[0101] The network index to which the trajectory point (x, y) belongs is:

[0102]

[0103] The grid size can be adjusted according to specific application requirements. If more sophisticated spatial segmentation and management are required, small grids are used. For macro analysis and large-scale data processing, larger grids are more suitable. It is suitable for everything from micro traffic management to macro urban planning. At the same time, grid slicing facilitates real-time updating and correction of vehicle trajectory data. New data can be quickly located in the corresponding grid unit for updating, which improves the dynamic adaptability of data.

[0104] S303, outlier correction;

[0105] An adaptive outlier screening algorithm based on sample entropy and local outlier factor is used to identify outliers in the trajectory. The outliers are checked and corrected according to speed, time, mileage and current position so that the entire trajectory conforms to the laws of kinematics. The sensitivity and accuracy of outlier detection and correction are improved through adaptive screening technology. The initial window size is set to L1, and the trajectory outliers are detected by sliding with the minimum allowable calculation amount. The sliding distance is T, and a threshold is set for the deviation characteristics of the outliers. The window first slides through the entire trajectory sequence with the initial size and T distance, locates the segment ST that exceeds the threshold and extracts it. The window is slid again based on ST. At this time, the initial accurate detection of outliers can be achieved through the principle of adaptive correction of window size. Subsequently, the local outlier factor algorithm is used to adaptively calculate the degree of outlier of the data point in the local area based on the preliminary screening results, considering the relationship between the data point and the overall data, and based on the density of the neighboring points around the data point. The main steps are as follows:

[0106] First, find the k nearest points of the target point d and calculate their reachability distance RD (reachability distance). The definition of reachability distance RD is as follows:

[0107] RD=max{k dis ,d dis (d,d j )}

[0108] In the formula, k dis is the distance from point d to the kth nearest neighbor; d dis (d,d j ) is point d to d j distance, j∈{1,2,…,k}.

[0109] Secondly, the vehicle trajectory is traversed and the inverse RD mean (LRD) of each point is calculated in turn and stored in the set R1; then, the trajectory is traversed again and the LOF (LOF is the abbreviation of Local Outlier Factor, which refers to the local anomaly factor. The present invention determines the outlier point based on the density of the data point, and assigns an outlier factor that depends on the neighborhood density to each data point to determine whether the data point is an outlier point) of each point is calculated in combination with R1 and stored in the set L1; the LOF corresponding to point d is expressed as follows:

[0110]

[0111] Finally, L1 is judged and screened according to the threshold c to obtain the final target outlier point set. Set the vehicle trajectory point set as {(x i ,y i ,t i )}, then the outlier is (xout ,y out ,t out ), the final target outlier point can be corrected by using the two outliers before and after:

[0112]

[0113] S304, generating trajectory information;

[0114] The vehicle trajectory points corrected in sub-step S303 are merged with the normal point set, and the generated trajectory information can be expressed as {(x j ,y j ,attribute j )}, where attributes j Including associated road network data, traffic element information and regional POI information. The cloud merges the corrected vehicle trajectory points with the normal point set to generate complete vehicle trajectory information, and synchronizes the results back to the vehicle for real-time update and feedback. Using incremental transmission technology, the data transmission formula is:

[0115]

[0116] Where, T incremental is the data transmission time, D is the total data transmission volume, t1 and t2 are the target synchronization times, and size(d) is the size of a single data record. After the receiver receives ΔD(t1, t2), it updates the local data set D(t1) to the latest state D(t2). The data relationship is as follows:

[0117]

[0118] Where D local For local datasets.

[0119] At the same time, the cloud analyzes and models historical trajectory data, generates predicted trajectories, and provides a more intelligent trajectory generation solution. The high-precision and real-time trajectory information generation can effectively improve the safety and efficiency of vehicle driving, support the construction and optimization of intelligent transportation systems, and improve the overall traffic management level.

Claims

1. A vehicle-cloud integrated vehicle trajectory data adaptive processing method, characterized by: The steps include: S1, data collection: on the vehicle side, a data collection device is used to collect vehicle trajectory data, the collected vehicle trajectory data includes the spatial coordinates and timestamp of the vehicle, and the collected vehicle trajectory data is transmitted to the vehicle's computing device through the vehicle's internal data bus or wireless communication; S2, data preprocessing: the vehicle trajectory data is preliminarily processed by the vehicle's computing device, including the following sub-steps: S201, redundant merging: setting a time window for redundant data, and merging vehicle trajectory points whose time difference is less than the time window; S202, null value filling; using adaptive window technology, dynamically adjusting the sampling window according to the vehicle trajectory data density; S203, data upload: uploading the pre-processed data to the cloud via a wireless network; S3, data association; includes the following sub-steps: S301, spatial association; using geographic spatial index to realize spatial data retrieval and management, spatially associating the redundant merged and null-filled vehicle trajectory points with the road network data, traffic element information and regional POI information; S302, data slicing; Slice the geospatial data according to the preset grid size to achieve grid management of spatial data; S303, outlier correction: using an adaptive outlier screening algorithm based on sample entropy and local outlier factor to identify outliers in the vehicle trajectory; S304, generating vehicle trajectory information: merging the corrected vehicle trajectory points with the normal point set to generate vehicle trajectory information.

2. The vehicle trajectory data adaptive processing method of vehicle-cloud integration according to claim 1, characterized in that: The data acquisition equipment includes video monitoring, ground sensing coils, wireless radio frequency identification, automotive electronic identification, and on-board diagnostic system equipment and global positioning system equipment installed in smart vehicles.

3. The vehicle trajectory data adaptive processing method of vehicle-cloud integration as claimed in claim 1, characterized in that: When executing S201, the vehicle trajectory data is set to: {(x i ,y i ,t i )},x i and i represents the spatial coordinates of the vehicle, t i Represents the timestamp, sets a time window Δt for redundant data, merges the vehicle trajectory data points with a time difference less than Δt, and the merged vehicle trajectory point position is the weighted average of all vehicle trajectory data points within the time window Δt: In the formula, w i is the weight, w i The value is set to 1 or according to the reliability of the data point.

4. The vehicle trajectory data adaptive processing method of vehicle-cloud integration as claimed in claim 3, characterized in that: For special working conditions, encrypted data points are used. Based on the concept of entropy, the complexity of vehicle trajectory data is calculated. For the two-dimensional trajectory coordinates, vector difference is used to perform sample entropy analysis, describe the complexity of instantaneous changes in the trajectory, capture the nonlinear characteristics in the signal, and set the threshold to adaptively and dynamically adjust the window size and fill in empty values ​​by judging the degree of change of sample entropy within a certain range.

5. The vehicle trajectory data adaptive processing method of vehicle-cloud integration as claimed in claim 4, characterized in that: The following steps are also included: First, a sampling window with an initial size of L is set, which slides across the vehicle trajectory sequence and calculates the sample entropy of the vehicle trajectory vector in each window segment; Secondly, a threshold is set and the sample entropy of adjacent segments is subtracted in pairs; the segments whose subtraction results exceed the threshold are regarded as segments with drastic changes in the curvature of the vehicle trajectory; Next, obtain the maximum length Lmax that meets the threshold judgment condition, and calculate the sample entropy mean of the vehicle trajectory corresponding to 1-Lmax; Next, sort the sample entropy means from small to large, draw a scatter plot and connect the minimum and maximum values ​​to form a straight line; Finally, the distance l from each point to the straight line is calculated in turn, and the horizontal coordinate corresponding to the maximum distance lmax is the optimal window length L.

6. The vehicle trajectory data adaptive processing method of vehicle-cloud integration as claimed in claim 5, characterized in that: For a given vehicle trajectory sequence B = {b1, b2, K, b n }, the sample entropy is calculated as follows: In the formula, the integer 0≤m≤N represents the length of the sequence to be compared, the real number r>0 represents the tolerance for accepting a match, and C m (r) represents the empirical probability of two sequences matching m points, which is obtained by the following formula: In the formula, Obtained by the following formula: In the formula, Represents any vector x m (j) in x m (i) The probability within the range r, NumT(x m (i),x m (j)) represents x m (i) and x m (j) is less than the number of times r, x m (i) and x m (j) is the m-dimensional vector reconstructed by B.

7. The vehicle trajectory data adaptive processing method of vehicle-cloud integration as claimed in claim 4, characterized in that: Use linear interpolation to fill in the empty values. Suppose two known vehicle trajectory points are (x i-1 ,y i-1 ,t i-1 ) and (x i+1 ,y i+1 ,t i+1 ), the missing point in the middle is (x i ,y i ,t i ),but: After obtaining L, the time window slides a certain step size and begins to adaptively process the next vehicle trajectory sequence.

8. The vehicle trajectory data adaptive processing method of vehicle-cloud integration as claimed in claim 1, characterized in that: When executing S302, the entire geographic space is divided into n×n grids, each grid has a size of Δx×Δy, and the network index to which the vehicle trajectory point (x, y) belongs is: When executing S303, the degree of outlier of the data point in the local area is adaptively calculated according to the density of neighboring points around the data point, including the following steps: First, find the k nearest points of the target point d and calculate their reachable distance RD. The definition of reachable distance RD is as follows: RD=max{k dis ,d dis (d,d j )}, In the formula, k dis is the distance from point d to the kth nearest neighbor; d dis (d,d j ) is point d to d j The distance, j∈{1,2,…,k}; Secondly, traverse the vehicle trajectory and calculate the RD mean reciprocal (LRD) of each point in turn and store it in set R1; Next, traverse the vehicle trajectory again and calculate the LOF of each point in combination with R1 and store it in the set L1 to obtain: Finally, L1 is judged and screened according to the threshold c to obtain the final target outlier point set.

9. The vehicle trajectory data adaptive processing method of vehicle-cloud integration as claimed in claim 1, characterized in that: Set the vehicle trajectory point set to {(x i ,y i ,t i )}, the outlier is (x out ,y out ,t out ), the final target outlier point can be corrected by using the two outliers before and after, and the relationship is as follows:

10. The vehicle trajectory data adaptive processing method of vehicle-cloud integration according to any one of claims 1 to 9, characterized in that: The cloud analyzes and models historical trajectory data to generate predicted trajectories.

Citation Information

Patent Citations

  • Vehicle track data processing method

    CN114724362A

Cited By

  • Obstacle avoidance path optimization method and system for airport unmanned inspection equipment

    CN122835406A