Illegal operation vehicle identification system and method based on spatio-temporal graph neural network

Abstract

The present invention relates to the field of intelligent transportation technology, and particularly to an illegal operation vehicle identification system and method based on a spatio-temporal graph neural network, which includes a data interaction module, a preprocessing module, a spatio-temporal graph construction module, a topology-enhanced hierarchical spatio-temporal attention network, a traffic feature extraction module, a data fusion module, and an intelligent decision-making module. The data interaction module connects to multiple platforms to obtain various types of data. The preprocessing module standardizes the format and encodes it. The spatio-temporal graph construction module constructs a spatio-temporal dynamic relationship graph containing multiple nodes. The topology-enhanced hierarchical spatio-temporal attention network extracts spatio-temporal joint feature vectors. The traffic feature extraction module extracts travel frequency and trip coherence features based on checkpoint data and calculates the probability of illegal operation. The data fusion module fuses the features to extract candidate illegal operation vehicles to form a vehicle library. The intelligent decision-making module calculates the feature weight coefficients and screens and determines illegal operation vehicles based on the entropy value method, greatly improving the identification coverage rate of illegal operation vehicles.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent transportation, and particularly to an illegal operation vehicle identification system and method based on a spatio-temporal graph neural network. Background Art

[0002] With the development of the sharing economy, the number of illegal operation vehicles has been increasing day by day. These vehicles do not have legitimate operation qualifications, which not only disrupt the normal order of the taxi market, but also pose a hidden danger to the safety of passengers. Traditional methods for identifying illegal operation vehicles mainly rely on manual inspections, which are inefficient and have limited coverage.

[0003] Existing technical solutions mainly include vehicle identification based on bayonet cameras, behavior analysis based on GPS trajectories, and post-event processing based on passenger reports, etc. These methods usually have the following problems: First, the information from a single data source is limited, making it difficult to comprehensively capture the behavioral characteristics of illegal operation vehicles; Second, traditional feature extraction methods are difficult to effectively handle the spatio-temporal correlation of vehicle behaviors, resulting in low identification accuracy; Third, there is a lack of an effective evidence storage and management mechanism, making it difficult to obtain law enforcement evidence.

[0004] In addition, existing methods usually adopt simple data fusion strategies and cannot make full use of the complementary characteristics of multi-source heterogeneous data; during the feature extraction process, ordinary neural networks are difficult to capture complex spatio-temporal structure information; in the decision-making stage, a single threshold is difficult to adapt to the differentiated requirements of different cities and scenarios. These problems seriously limit the actual application effect of the illegal operation vehicle identification system. Summary of the Invention

[0005] The purpose of the present invention is to provide an illegal operation vehicle identification system and method based on a spatio-temporal graph neural network, which can achieve high-precision identification of illegal operation vehicles by introducing a topology-enhanced hierarchical spatio-temporal attention network, and effectively solve the problems of incomplete feature extraction, low identification accuracy, and imperfect evidence management in the prior art.

[0006] The present invention proposes an illegal operation vehicle identification system based on a spatio-temporal graph neural network, including:

[0007] A data interaction module, used to connect to a mobile payment management platform, an urban traffic control platform, and a blockchain evidence storage platform, and obtain bayonet capture data, mobile payment platform trip data, urban traffic control platform data, and passenger APP operation behavior data;

[0008] A data preprocessing module, connected to the data interaction module, used to unify the formats of the bayonet capture data, the mobile payment platform trip data, the urban traffic control platform data, and the passenger APP operation behavior data, and perform encoding standardization on each dimension of data;

[0009] A spatio-temporal graph construction module, connected to the data preprocessing module, for constructing a spatio-temporal dynamic relationship graph including driver nodes, checkpoint nodes, vehicle nodes, vehicle type nodes, trip nodes, and parking point nodes based on the encoded and standardized data;

[0010] A topological enhanced hierarchical spatio-temporal attention network, connected to the spatio-temporal graph construction module, including a topological space mapping layer, a probability attention weighting layer, and a spatio-temporal feature fusion layer, for extracting spatio-temporal joint feature vectors from the spatio-temporal dynamic relationship graph;

[0011] A traffic feature extraction module, connected to the data preprocessing module, for extracting travel frequency and trip coherence features based on the time-space trajectory of the checkpoint capture data, and calculating the probability of illegal operation;

[0012] A data fusion module, connected to the topological enhanced hierarchical spatio-temporal attention network and the traffic feature extraction module, for fusing the spatio-temporal joint feature vectors, the travel frequency, and the trip coherence features, extracting candidate illegal operation vehicles, and forming an illegal operation vehicle library;

[0013] An intelligent decision-making module, connected to the data fusion module, for calculating the weight coefficients of each dimension feature, screening the vehicles in the illegal operation vehicle library based on the entropy value method, and determining the illegal operation vehicles.

[0014] Preferably, the topological space mapping layer is used for:

[0015] Representing the urban road network as a graph structure, where intersections are nodes and roads are edges;

[0016] Assigning weights to each edge of the graph structure, where the weights represent road travel time or travel difficulty;

[0017] Constructing a road adjacency matrix to represent the connectivity of the road network;

[0018] Mapping the vehicle GPS trajectory point sequence to the nearest road edge;

[0019] Calculating the path representation of the trajectory in the topological space;

[0020] Extracting the topological invariant features of the path;

[0021] Calculating the topological persistence index of the trajectory to represent the trajectory structure stability;

[0022] Identifying the topological features in the trajectory;

[0023] Generating a trajectory topological signature.

[0024] Preferably, the probability attention weighting layer is used for:

[0025] Decompose the trajectory into sub-trajectories of different time scales;

[0026] Extract features for each time scale to capture behavioral patterns at different time granularities;

[0027] Construct a multi-scale feature matrix to represent the feature distribution at different time scales;

[0028] Construct a conditional probability model to quantify the importance of features for the judgment of illegal operation;

[0029] Calculate the information gain of each feature as the basis for attention weights;

[0030] Design three attention heads for time, space, and spatio-temporal joint;

[0031] Highlight abnormal behavior features through the attention mechanism to improve detection sensitivity.

[0032] Preferably, the spatio-temporal feature fusion layer is used for:

[0033] Analyze the complementarity and redundancy of features from different sources;

[0034] Remove highly correlated features and retain complementary information;

[0035] Construct a feature dependency graph to guide the feature fusion strategy;

[0036] Adjust the fusion weights according to factors such as city characteristics, time periods, and weather;

[0037] Construct joint features representing the starting area, travel path, passing areas, etc.;

[0038] Retain the high-order correlation between features;

[0039] Generate the final spatio-temporal joint feature vector.

[0040] Preferably, the data preprocessing module is further used for:

[0041] Standardize the time information to unify the time representation format of data from different sources;

[0042] Standardize the space information and map the location information to a unified geographic coordinate system;

[0043] Standardize the vehicle information to unify the vehicle identification representation from different sources;

[0044] Perform outlier detection and processing to identify and process data that significantly deviates from the normal range;

[0045] Perform missing value imputation and make reasonable inferences about the missing data based on spatio-temporal correlation;

[0046] Perform data consistency verification to check whether the same information from different sources is consistent.

[0047] Preferably, the space-time graph construction module is further used for:

[0048] Set driving habits, driving time and other attributes for the driver node;

[0049] Set geographic location, passing frequency and other attributes for the checkpoint node;

[0050] Set attributes such as vehicle model and license plate color for vehicle nodes;

[0051] Set attributes such as function type and regional characteristics for location nodes;

[0052] Construct vehicle-checkpoint edges to represent the time and frequency of vehicles passing through the checkpoints;

[0053] Construct vehicle-location edges to represent the vehicle’s stay time and frequency at the location;

[0054] Construct driver-vehicle edges to represent driving association strength and time distribution;

[0055] Construct location-location edges to represent the travel frequency and travel time between locations;

[0056] Implement dynamic graph updates and update the graph structure when new data arrives.

[0057] Preferably, the traffic feature extraction module is further used for:

[0058] Calculate the travel frequency and count the number of trips in different time windows;

[0059] Analyze the continuity of the trip and evaluate the rationality of the time and space logic between consecutive trips;

[0060] Extract parking patterns, identify regular parking spots and duration distribution;

[0061] Calculate the Gini coefficient of travel trajectories to quantify the inequality of travel distances;

[0062] Setting a Gini coefficient threshold to determine professional driver status;

[0063] Divide the urban area into several sectors;

[0064] Use the boundary checkpoints of each sector to construct a spatiotemporal feature set;

[0065] According to the feature similarity of the feature set, the boundary checkpoints are clustered into several clusters;

[0066] Calculate the average dwell time at checkpoints within the cluster and the probability of illegal operation.

[0067] Preferably, the system further includes:

[0068] A passenger APP feature extraction module, connected to the data interaction module, for extracting user rating patterns, search behaviors, and payment records from passenger APP operation behavior data, and calculating the possibility of illegal operation;

[0069] The data fusion module is also used to fuse the possibility of illegal operation.

[0070] Preferably, the system further includes:

[0071] A blockchain evidence storage module, connected to the intelligent decision-making module, for receiving and encrypting and storing the recognition results of illegal operation vehicles, including information such as MAC address, usage time, license plate, driving route, etc.;

[0072] Perform block storage according to a preset time period;

[0073] Provide verifiable law enforcement evidence.

[0074] An illegal operation vehicle recognition method based on a spatio-temporal graph neural network, adopting the described system, is characterized by including:

[0075] Connect to the mobile payment management platform, urban traffic control platform, and blockchain evidence storage platform through the data interaction module to obtain bayonet capture data, mobile payment platform trip data, urban traffic control platform data, and passenger APP operation behavior data;

[0076] Unify the formats of the bayonet capture data, the mobile payment platform trip data, the urban traffic control platform data, and the passenger APP operation behavior data through the data preprocessing module, and perform encoding standardization on the data of each dimension;

[0077] Construct a spatio-temporal dynamic relationship graph including driver nodes, bayonet nodes, vehicle nodes, vehicle type nodes, trip nodes, and parking point nodes based on the encoded and standardized data through the spatio-temporal graph construction module;

[0078] Extract spatio-temporal joint feature vectors from the spatio-temporal dynamic relationship graph through a topological enhanced hierarchical spatio-temporal attention network, and the topological enhanced hierarchical spatio-temporal attention network includes a topological space mapping layer, a probability attention weighting layer, and a spatio-temporal feature fusion layer;

[0079] Extract the travel frequency and trip coherence features based on the time-space trajectory of the bayonet capture data through the traffic feature extraction module, and calculate the probability of illegal operation;

[0080] The spatio-temporal joint feature vector, the travel frequency, and the trip coherence feature are fused through the data fusion module to extract candidate illegal operation vehicles and form an illegal operation vehicle library;

[0081] The weight coefficients of the features in each dimension are calculated through the intelligent decision-making module, and the vehicles in the illegal operation vehicle library are screened based on the entropy method to determine illegal operation vehicles.

[0082] The present invention has the following beneficial effects:

[0083] 1. Through the multi-source heterogeneous data fusion mechanism, integrating the data captured by the checkpoint, the trip data of the mobile payment platform, the data of the urban traffic control platform, and the operation behavior data of the passenger APP, more comprehensive vehicle behavior feature capture is realized, and the recognition coverage rate of illegal operation vehicles is greatly improved.

[0084] 2. An innovative topological enhanced hierarchical spatio-temporal attention network is adopted, and the essential structural features of vehicle trajectories are extracted through a three-layer progressive structure (topological space mapping layer, probability attention weighting layer, and spatio-temporal feature fusion layer). Compared with traditional methods, the recognition accuracy is increased by about 30%, and the misjudgment rate is reduced by about 50%.

[0085] 3. The blockchain evidence storage technology is introduced to ensure the reliability and immutability of law enforcement evidence, and the success rate of illegal operation law enforcement is greatly improved.

[0086] 4. The system has an adaptive characteristic, and can automatically adjust the model parameters according to different city characteristics, seasonal changes, and weather factors, significantly improving the robustness and adaptability of the system.

[0087] 5. Based on the multi-dimensional feature weight optimization mechanism of the entropy method, the automatic evaluation and dynamic adjustment of feature importance are realized, avoiding the subjectivity of manually setting weights and improving the objective fairness of the system. Brief Description of the Drawings

[0088] Figure 1 It is the overall architecture diagram of the illegal operation vehicle recognition system based on the spatio-temporal graph neural network of the present invention;

[0089] Figure 2 It is the structural schematic diagram of the topological enhanced hierarchical spatio-temporal attention network of the present invention;

[0090] Figure 3 It is the working flow chart of the spatio-temporal graph construction module of the present invention;

[0091] Figure 4 It is the feature fusion schematic diagram of the data fusion module of the present invention;

[0092] Figure 5 It is the decision flow chart of the intelligent decision-making module of the present invention;

[0093] Figure 6 This is the flowchart of the illegal operation vehicle identification method of the present invention. Specific embodiments

[0094] Please refer to the appendix Figure 1-6 In combination with the accompanying drawings and specific embodiments, the present invention will be further described in detail below.

[0095] As Figure 1 shown, the illegal operation vehicle identification system based on the spatio-temporal graph neural network provided by the present invention includes a data interaction module 10, a data preprocessing module 20, a spatio-temporal graph construction module 30, a topological enhanced hierarchical spatio-temporal attention network 40, a traffic feature extraction module 50, a data fusion module 60, an intelligent decision-making module 70, a passenger APP feature extraction module 80, and a blockchain evidence storage module 90.

[0096] The data interaction module 10 is used to connect to the mobile payment management platform, the urban traffic control platform, and the blockchain evidence storage platform, and obtain bayonet capture data, mobile payment platform itinerary data, urban traffic control platform data, and passenger APP operation behavior data. Preferably, the data interaction module 10 uses a standard API interface to exchange data with each platform, and supports two modes of real-time data stream and batch data acquisition. In an embodiment of the present invention, the data interaction module 10 obtains itinerary data from the mobile payment platform every 5 minutes and obtains bayonet capture data from the urban traffic control platform every 1 minute to ensure the timeliness of the data.

[0097] The data preprocessing module 20 is connected to the data interaction module 10, and is used to unify the formats of the bayonet capture data, mobile payment platform itinerary data, urban traffic control platform data, and passenger APP operation behavior data, and encode and standardize the data of each dimension. Preferably, this module uniformly formats the time information and adopts the standard time format of YYYY-MM-DD HH:MM:SS; performs coordinate system conversion on the spatial information and uniformly adopts the WGS84 coordinate system; standardizes the license plate information, removes special characters and spaces, and facilitates subsequent matching and analysis. In addition, the data preprocessing module 20 is also responsible for identifying and processing abnormal data, such as GPS coordinates that deviate significantly from the reasonable range, records with incorrect timestamps, etc.

[0098] The spatio-temporal graph construction module 30, connected to the data preprocessing module 20, is used to construct a spatio-temporal dynamic relationship graph containing driver nodes, checkpoint nodes, vehicle nodes, vehicle type nodes, trip nodes, and parking point nodes based on the encoded and standardized data. In the spatio-temporal dynamic relationship graph, nodes represent entities, and edges represent the relationships between entities. Specifically, vehicle-checkpoint edges are constructed to represent the time and frequency of vehicles passing through checkpoints; vehicle-location edges are constructed to represent the residence time and frequency of vehicles at locations; driver-vehicle edges are constructed to represent the driving association strength and time distribution; location-location edges are constructed to represent the trip frequency and passing time between locations. At the same time, the spatio-temporal graph construction module 30 also supports dynamic graph updates. When new data arrives, the graph structure is updated in a timely manner to ensure the timeliness of the graph.

[0099] The topology-enhanced hierarchical spatio-temporal attention network 40, connected to the spatio-temporal graph construction module 30, includes a topological space mapping layer 41, a probability attention weighting layer 42, and a spatio-temporal feature fusion layer 43, and is used to extract spatio-temporal joint feature vectors from the spatio-temporal dynamic relationship graph. Preferably, the spatio-temporal joint feature vector is a vector with a dimension of 128, containing the spatio-temporal behavior characteristics of the vehicle, and is used for subsequent illegal operation judgment.

[0100] The traffic feature extraction module 50, connected to the data preprocessing module 20, is used to extract travel frequency and trip coherence features based on the time-space trajectory of checkpoint capture data and calculate the probability of illegal operation. Specifically, this module counts the number of trips of vehicles within different time windows, such as the travel frequency distribution within 24 hours, on weekdays, weekends, etc.; analyzes the spatio-temporal logical rationality between consecutive trips to determine whether there are abnormal consecutive trips; identifies the regular parking points and parking duration distribution of vehicles to determine whether the vehicle has professional operation characteristics.

[0101] The data fusion module 60, connected to the topology-enhanced hierarchical spatio-temporal attention network 40 and the traffic feature extraction module 50, is used to fuse the spatio-temporal joint feature vectors, travel frequency, and trip coherence features, extract candidate illegal operation vehicles, and form an illegal operation vehicle library. Preferably, the data fusion module 60 uses a weighted fusion method, assigns weights according to the discrimination ability of different features, generates a comprehensive score, and screens out candidate illegal operation vehicles according to the score.

[0102] The intelligent decision-making module 70, which is connected to the data fusion module 60, is used to calculate the weight coefficients of the features in each dimension, screen the vehicles in the illegal operation vehicle library based on the entropy method, and determine the illegal operation vehicles. In the embodiment of the present invention, the entropy method automatically assigns weights according to the information amount of the features, avoiding the deviation of subjectively setting weights. Preferably, the intelligent decision-making module 70 sets the comprehensive scoring threshold to 0.75 (with a full score of 1), and the vehicles with a score higher than this threshold are determined as illegal operation vehicles. The setting of this threshold is based on a large number of historical case analyses and expert experiences, which can control the misjudgment rate while ensuring the recognition rate.

[0103] The topological enhanced hierarchical spatio-temporal attention network 40 of the present invention is the core innovation of the system, including three levels: a topological space mapping layer 41, a probability attention weighting layer 42, and a spatio-temporal feature fusion layer 43, which realizes high-quality spatio-temporal feature extraction in a progressive manner.

[0104] The implementation method of the topological space mapping layer 41 is as follows:

[0105] First, represent the urban road network as a graph structure , where is the set of intersection nodes, is the set of road edges. For each road , assign a weight , representing the travel time or travel difficulty of the road. In actual implementation, the weight can be calculated based on factors such as road length, travel speed, and congestion situation. For example:

[0106] ,

[0107] where, is the road length (unit: meter), is the average travel speed of the road (unit: meter / second), is the road congestion index (value range: between 0 and 1), is the congestion impact factor, used to adjust the impact of congestion on travel time, and generally takes a value of 0.5.

[0108] Then, construct the road adjacency matrix , representing the connectivity of the road network, where the element is defined as:

[0109] ,

[0110] where, represents the element in the matrix at the th row and the th column, and the value of 1 indicates that the node and the node There is a road connection between them. A value of 0 indicates no connection. The matrix has a dimension of , which is the square of the number of nodes.

[0111] Then, map the vehicle GPS trajectory point sequence to the nearest road edge to obtain the road segment sequence . To improve the mapping accuracy, use the hidden Markov model for trajectory matching, considering the road network topology constraints to ensure the continuity of the mapping result.

[0112] Among them, represents the -th point in the GPS trajectory, including longitude, latitude, and timestamp information, is the total number of GPS points; represents the -th road segment in the mapped road segment sequence, corresponding to an edge in the road network, is the length of the road segment sequence. Usually , because multiple adjacent GPS points may be mapped to the same road segment. Based on the mapped road segment sequence, calculate the topological persistence index of the trajectory, which represents the stability of the trajectory structure:

[0113] ,

[0114] Among them, is the persistence value of the -th topological feature in the trajectory, quantifying the stability of this feature during the trajectory deformation process. The larger the value, the more stable it is; is the number of topological features considered, generally taking , corresponding to 0-dimensional, 1-dimensional, and 2-dimensional topological features (representing connected components, loops, and holes respectively).

[0115] By analyzing the topological structure of the trajectory, identify specific patterns, such as illegal operation features like detours and loops. Finally, generate the trajectory topological signature , as the basis for distinguishing different behavior patterns:

[0116] ,

[0117] Among them, is the aforementioned topological persistence index, is the number of connected components of the trajectory, representing the number of disconnected parts in the trajectory, is the number of loops in the trajectory, representing the number of closed loops in the trajectory. is a three-dimensional vector. In practical applications, it can be further extended to a high-dimensional vector, including more topological features.

[0118] The implementation of the probability attention weighted layer 42 is as follows: First, decompose the trajectory into a set of sub-trajectories at different time scales , corresponding to hourly, daily, and weekly levels respectively. Extract features for each time scale to construct a multi-scale feature matrix :

[0119] ,

[0120] where represents the feature vector extracted from the trajectory at time scale , representing the hourly, daily, and weekly time scales respectively; is the multi-scale feature matrix with a dimension of , where is the dimension of a single feature vector (such as 128), and 3 represents the three time scales.

[0121] Then, construct a conditional probability model to quantify the importance of features for illegal operation judgment,

[0122] ,

[0123] where is a binary variable indicating whether it is an illegal operation (1 means yes, 0 means no), is the th feature, is the weight parameter of this feature, learned from the training data; represents the probability range of the vehicle being an illegal operation under the condition of the given feature ranging from [0, 1]. Calculate the information gain of each feature as the basis for the attention weight:

[0124] ,

[0125] where is the entropy of the target variable , representing a measure of uncertainty, and the calculation formula is is the conditional entropy of the target variable under the given feature , and the calculation formula is ;

[0126] represents the information gain provided by the feature , and the larger the value, the greater the contribution of this feature to the judgment .

[0127] Design three attention heads for time, space, and spatio-temporal joint, respectively focusing on features in different dimensions:

[0128] The time attention head focuses on time regularity features, such as peak-hour activity frequency, night-time activity patterns, etc.;

[0129] The space attention head focuses on space distribution features, such as frequently visited locations, activity ranges, etc.;

[0130] The spatio-temporal joint attention head focuses on spatio-temporal interaction features, such as spatial distribution during a specific time period, etc.

[0131] Finally, generate a weighted feature vector through the attention mechanism:

[0132] ,

[0133] where, is the attention weight of the th feature, satisfying , is the total number of features; is the th feature; is the weighted feature vector, with the same dimension as a single . The attention weight is calculated as:

[0134] ,

[0135] where, represents the natural exponential function, is the aforementioned information gain; this formula implements softmax normalization to ensure that the sum of all weights is 1.

[0136] The spatio-temporal feature fusion layer 43 is implemented as follows: First, analyze the complementarity and redundancy of features from different sources, and calculate the correlation matrix of the features:

[0137] ,

[0138] where, is the covariance of features and , and the calculation formula is , represents the expectation operation, and are the means of features and respectively; and are features and The standard deviation of Representation characteristics and The Pearson correlation coefficient ranges from [-1,1]. The larger the absolute value, the stronger the correlation.

[0139] Remove highly correlated features and retain complementary information. Generally speaking, when When considering removing features with low information content to construct feature dependency graphs ,in is a set of feature nodes, each node corresponds to a feature is the set of dependency edges between features, and the edge weight is the absolute value of the correlation coefficient.

[0140] in, Indicates the absolute value of the correlation coefficient; the threshold of 0.85 is set based on experience and represents the judgment standard of high correlation. The number of nodes is equal to the number of features, and the number of edges depends on the correlation distribution between features. The fusion weights are adjusted according to city characteristics, time periods, weather and other factors. For example, for the core areas of large cities, the spatial attention weight is appropriately increased; for peak hours on weekdays, the temporal attention weight is increased; for severe weather conditions, the weights of the corresponding features are appropriately reduced.

[0141] Finally, a joint feature vector representing the starting area, travel path, passing area, etc. is constructed :

[0142] ,

[0143] in, , and They are time, space, and spatiotemporal joint feature vectors, respectively, and their dimensions may be different; Concat represents a vector concatenation operation, which connects the three vectors end to end to form a longer vector; is the final joint feature vector, whose dimension is the sum of the dimensions of the three input vectors. In order to retain the high-order correlation between features, the tensor decomposition method can be further used to extract the potential feature patterns.

[0144] The detailed implementation of the data preprocessing module 20 is as follows:

[0145] Standardize time information and unify the time representation format of data from different sources. Specifically, convert all time data into the YYYY-MM-DDHH:MM:SS format and unify it into the UTC+8 time zone (Beijing time). For data with missing time zone information, infer the possible time zone based on the geographical location.

[0146] Standardize the spatial information and map the location information to a unified geographic coordinate system. Preferably, the WGS84 coordinate system is used as the standard coordinate system, and for data using other coordinate systems (such as GCJ-02, BD-09, etc.), coordinate conversion is performed. The specific conversion formulas are as follows:

[0147] Conversion from GCJ-02 to WGS84:

[0148] ,

[0149] ,

[0150] where and are the latitude and longitude (in degrees) in the GCJ-02 coordinate system, and are the latitude and longitude (in degrees) in the WGS84 coordinate system; and respectively represent the differences in latitude and longitude between the two coordinate systems. The true WGS84 coordinates are approximated through an iterative method, and generally, 10 iterations can achieve centimeter-level accuracy.

[0151] Standardize the vehicle information and unify the representation of vehicle identifiers from different sources. The license plate number is unified into the format of uppercase letters plus numbers, removing special characters and spaces. For new energy vehicle license plates, their special identifiers are retained. Perform outlier detection and processing to identify and process data that significantly deviates from the normal range. Use the Z-score method to identify outliers:

[0152] ,

[0153] where is the value to be detected, is the mean of this feature, is the standard deviation; is the standardized value, also known as the Z-score. When (i.e., the absolute value is greater than 3), it is regarded as an outlier. This threshold is based on the empirical rule in statistics. Under the assumption of a normal distribution, approximately 99.7% of the data points fall within 3 standard deviations of the mean . For outliers, methods such as replacing with adjacent values, replacing with the mean, or directly removing can be used for processing.

[0154] Reasonably supplement the missing data and infer the missing data based on spatio-temporal correlation. Use the K-nearest neighbor (KNN) method for missing value filling:

[0155] ,

[0156] Among them, is the missing value to be filled, is the corresponding eigenvalue of the th sample that is most similar to the missing value sample, is the number of neighboring samples considered, generally taking 3 - 5. The similarity is calculated based on known features, usually using Euclidean distance or Manhattan distance metrics. Data consistency verification is performed to check whether the same information from different sources is consistent. For example, compare the time and location information in bayonet capture data and travel data. When the degree of inconsistency exceeds the threshold, the data source with higher credibility is adopted. Specifically, define the consistency metric :

[0157] ,

[0158] Among them, and are the same information from two sources (such as location, time, etc.), is its difference metric (such as time difference, distance, etc.), is the maximum acceptable difference; The value range is [0, 1]. The closer it is to 1, the higher the consistency. When , it is regarded as inconsistent and needs to be processed.

[0159] The detailed implementation method of the traffic feature extraction module 50 is as follows:

[0160] The traffic feature extraction module 50 first calculates the travel frequency of the vehicle and counts the number of trips within different time windows. Specifically, it includes: the number of trips within 24 hours , the average number of trips on weekdays , the average number of trips on weekends , the number of trips during peak hours (7:00 - 9:00 and 17:00 - 19:00) , etc. Professional operating vehicles usually have a relatively high travel frequency, especially during peak hours.

[0161] Then, analyze the trip coherence and evaluate the spatio - temporal logical rationality between consecutive trips. Calculate the spatio - temporal connection coefficient :

[0162] ,

[0163] Among them, is the th trip end point and the th trip start point The distance between them (unit: meter), is the reasonable maximum driving speed (unit: m / s, for example, 60 km / h is equal to 16.67 m / s), and are the start time and end time (unit: s) of the th trip respectively, is the total number of trips; The ideal range of the value is [0, 1]. The closer it is to 1, the more reasonable the coherence is. If it is too small, it may indicate missed trips. If it is too large, it may indicate speeding.

[0164] Next, extract the parking mode features to identify regular parking points and duration distribution. Use the density-based clustering method (such as DBSCAN) to identify regular parking points:

[0165] ,

[0166] where is the set of parking points, and each point contains longitude and latitude coordinate information; is the clustering radius parameter (usually set to 100 m), indicating the range of the neighborhood; MinPts is the minimum number of points parameter (usually set to 3), indicating the minimum number of neighborhoods required to form a core point; Cluster is the clustering result, containing multiple clusters, and each cluster corresponds to a regular parking area. Calculate the parking duration distribution for each clustering to obtain the average parking duration , the longest parking duration and features such as the parking frequency Freq.

[0167] The traffic feature extraction module 50 also calculates the Gini coefficient of the travel trajectory to quantify the inequality of travel distances:

[0168] ,

[0169] where is the distance (unit: m) of the th trip, is the distance (unit: m) of the th trip, is the total number of trips, is the average travel distance (unit: m), calculated as is the Gini coefficient, and its value range is [0, 1]. The larger the value, the more uneven the travel distance distribution. The double summation calculates the sum of the absolute differences between all pairs of travel distances, and divides by for normalization. For professional operating vehicles The value is usually low because its travel distance distribution is relatively uniform. Preferably, the Gini coefficient threshold is set to 0.3, and vehicles with a value lower than this threshold are considered to have professional operation characteristics. In addition, the traffic feature extraction module 50 divides the urban area into several sectors. Based on the city center or major transportation hubs, the city is divided into 8 - 12 sectors. A spatio-temporal feature set is constructed using the boundary checkpoints of each sector, including the time distribution, frequency, etc. of vehicles entering and leaving the sector. The boundary checkpoints are clustered according to feature similarity, and the average residence time of checkpoints within the cluster is calculated :

[0170] ,

[0171] where is the residence duration of a vehicle with a residence time of (unit: minute), is the number of vehicles with a residence time of , is the number of types of different residence durations, is the cluster number; is the average residence time within the cluster (unit: minute). The numerator represents the sum of the residence times of all vehicles, and the denominator represents the total number of vehicles. Dividing the two gives the average residence time. Based on the residence time distribution within the cluster, the probability of illegal operation is calculated :

[0172] ,

[0173] The probability index of illegal operation within the cluster . The larger the value, the more abnormal it is. Where is the residence time of vehicle within the cluster (unit: minute), is the average residence time within the cluster (unit: minute), and the calculation formula is , where is the total number of vehicles within the cluster ; represents the probability index of illegal operation of vehicle within the cluster . The larger the value, the more abnormal it is. This index is essentially the absolute value of the Z-score of the residence time, measuring the degree of deviation of the vehicle's residence time from the average level

[0174] The present invention further includes a passenger APP feature extraction module 80 and a blockchain evidence storage module 90. The passenger APP feature extraction module 80 is connected to the data interaction module 10 and is used to extract user scoring patterns, search behaviors, and payment records from the passenger APP operation behavior data to calculate the likelihood of illegal operation. Specifically, this module analyzes the scoring distribution of users for drivers. Abnormally high or consistent scores may imply abnormal operation behaviors; it analyzes users' search behaviors, including search frequency, content, and time distribution. Frequent searches for specific routes or searches at fixed times may indicate professional operation characteristics; it statistically analyzes the amount distribution, frequency, and time patterns of payment records to discover abnormal payment patterns. Based on these features, a likelihood score of illegal operation is calculated :

[0175] ,

[0176] wherein, , and are scores calculated based on the scoring pattern, search behavior, and payment record respectively, and are all normalized to the interval [0,1]; , and are the corresponding weights, satisfying , usually taking , indicating the relative importance of the three features. In addition to fusing the spatio-temporal joint feature vector, travel frequency, and trip coherence features, the data fusion module 60 also fuses the likelihood score of illegal operation calculated by the passenger APP feature extraction module 80 to form a more comprehensive feature representation and improve the accuracy of judgment.

[0177] The blockchain evidence storage module 90 is connected to the intelligent decision-making module 70 and is used to receive and encrypt and store the identification results of illegal operation vehicles, including information such as MAC address, usage time, license plate, and driving route. This module uses distributed ledger technology to ensure that the stored evidence cannot be tampered with and provides reliable technical support for subsequent law enforcement.

[0178] Specifically, the blockchain evidence storage module 90 generates and stores blocks according to a preset time period (such as every 24 hours). Each block contains evidence of illegal operation within a certain time period. The blocks are linked by hashes, and the hash value of each block is calculated as follows:

[0179] ,

[0180] wherein, is the hash value of the previous block, is the data content of the current block, is the timestamp, || represents the string concatenation operation, Denote a hash function (such as SHA-256); is the hash value of the current block and serves as the unique identifier of the block. This chained structure ensures the immutability of historical data. Any modification to historical blocks will cause the hash values of all subsequent blocks to change, facilitating detection.

[0181] Meanwhile, this module provides an evidence query interface, supporting law enforcement departments to query relevant evidence according to conditions such as license plate numbers and time periods, facilitating evidence collection and law enforcement. The query results include a complete evidence chain and provide a verification mechanism to ensure the authenticity and integrity of the evidence.

[0182] The method for identifying illegal operation vehicles based on spatio-temporal graph neural networks of the present invention includes the following steps:

[0183] Connect to the mobile payment management platform, urban traffic control platform, and blockchain evidence storage platform through the data interaction module 10 to obtain bayonet capture data, mobile payment platform itinerary data, urban traffic control platform data, and passenger APP operation behavior data. This step establishes a multi-source data acquisition channel and provides comprehensive data support for subsequent processing.

[0184] Unify the formats of the bayonet capture data, mobile payment platform itinerary data, urban traffic control platform data, and passenger APP operation behavior data through the data preprocessing module 20, and perform encoding standardization on the data of each dimension. This step ensures the consistency and comparability of data from different sources and lays a foundation for subsequent analysis.

[0185] Construct a spatio-temporal dynamic relationship graph containing driver nodes, bayonet nodes, vehicle nodes, vehicle type nodes, itinerary nodes, and parking point nodes based on the encoded and standardized data through the spatio-temporal graph construction module 30. This step organizes various entities and their relationships into a graph structure, facilitating the capture of complex spatio-temporal association patterns.

[0186] Extract spatio-temporal joint feature vectors from the spatio-temporal dynamic relationship graph through the topological enhanced hierarchical spatio-temporal attention network 40. This step is the core innovation of the method and realizes high-quality feature extraction through a three-layer progressive structure (topological space mapping layer 41, probabilistic attention weighting layer 42, and spatio-temporal feature fusion layer 43).

[0187] Extract the travel frequency and itinerary coherence features based on the time-space trajectory of the bayonet capture data through the traffic feature extraction module 50, and calculate the probability of illegal operation. This step analyzes vehicle features from the perspective of traffic behavior and identifies typical patterns of professional operation vehicles.

[0188] Fuse the spatio-temporal joint feature vectors, travel frequency, and itinerary coherence features through the data fusion module 60 to extract candidate illegal operation vehicles and form an illegal operation vehicle library. This step integrates multi-dimensional features and improves the comprehensiveness and accuracy of judgment.

[0189] The weight coefficients of the features in each dimension are calculated by the intelligent decision-making module 70, and the vehicles in the illegal operation vehicle library are screened based on the entropy value method to determine the illegal operation vehicles. This step realizes the final determination of illegal operation and outputs the recognition result.

[0190] The entire method process forms a complete closed-loop for identifying illegal operation vehicles. From data acquisition, preprocessing, feature extraction to the final decision-making, each link is closely connected and works collaboratively to achieve efficient and accurate identification of illegal operation vehicles.

[0191] The illegal operation vehicle identification system based on the spatio-temporal graph neural network of the present invention can be deployed on the server cluster of the urban traffic management center and connected to each data platform through standard interfaces. The system supports distributed deployment and can be flexibly expanded according to the city scale and data volume.

[0192] In a practical application case, the system is deployed on the traffic management platform of a first-tier city, processing approximately 5 million bayonet capture records and 3 million trip data every day. The system analyzes approximately 1 million active vehicles and identifies approximately 200 suspected illegal operation vehicles on average every day. After verification by the traffic law enforcement department, the recognition accuracy rate reaches over 85%, greatly improving the law enforcement efficiency. At the same time, the blockchain evidence storage function of the system provides reliable evidence for law enforcement, and the law enforcement success rate is increased by approximately 40%.

[0193] In addition, the system also has strong adaptability and can automatically adjust parameters according to the characteristics of different cities. For example, in tourist cities, the system will appropriately lower the judgment threshold around scenic spots to reduce misjudgment of tourist vehicles; in cities with severe traffic congestion, the system will adjust the evaluation standard for trip coherence to adapt to the special traffic environment.

[0194] In summary, the illegal operation vehicle identification system and method based on the spatio-temporal graph neural network provided by the present invention achieve high-precision identification of illegal operation vehicles through the innovative topological enhanced hierarchical spatio-temporal attention network, and have broad application prospects and significant social value.

[0195] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. An illegal operation vehicle identification system based on a spatio-temporal graph neural network, characterized in that, Including: A data interaction module, used to connect the mobile payment management platform, the urban traffic control platform and the blockchain evidence storage platform, and obtain bayonet capture data, mobile payment platform travel data, urban traffic control platform data and passenger APP operation behavior data; A data preprocessing module, connected to the data interaction module, used to unify the formats of the bayonet capture data, the mobile payment platform travel data, the urban traffic control platform data and the passenger APP operation behavior data, and perform encoding standardization on data of each dimension; A spatio-temporal graph construction module, connected to the data preprocessing module, used to construct a spatio-temporal dynamic relationship graph including driver nodes, bayonet nodes, vehicle nodes, vehicle type nodes, travel nodes and parking point nodes based on the encoded and standardized data; A topological enhanced hierarchical spatio-temporal attention network, connected to the spatio-temporal graph construction module, including a topological space mapping layer, a probability attention weighting layer and a spatio-temporal feature fusion layer, used to extract spatio-temporal joint feature vectors from the spatio-temporal dynamic relationship graph; A traffic feature extraction module, connected to the data preprocessing module, used to extract travel frequency and travel coherence features based on the time-space trajectory of the bayonet capture data, and calculate the probability of illegal operation; A data fusion module, connected to the topological enhanced hierarchical spatio-temporal attention network and the traffic feature extraction module, used to fuse the spatio-temporal joint feature vectors, the travel frequency and the travel coherence features, extract candidate illegal operation vehicles, and form an illegal operation vehicle library; An intelligent decision-making module, connected to the data fusion module, used to calculate the weight coefficients of each dimension feature, screen the vehicles in the illegal operation vehicle library based on the entropy value method, and determine illegal operation vehicles.

2. The illegal operation vehicle identification system based on the spatio-temporal graph neural network according to claim 1, characterized in that, The topological space mapping layer is used for: Representing the urban road network as a graph structure, where intersections are nodes and roads are edges; Assigning weights to each edge of the graph structure, and the weights represent road passing time or passing difficulty; Constructing a road adjacent matrix to represent the connectivity of the road network; Mapping the vehicle GPS trajectory point sequence to the nearest road edge; Calculating the path representation of the trajectory in the topological space; Extracting the topological invariant features of the path; Calculating the topological persistence index of the trajectory to represent the trajectory structure stability; Identifying the topological features in the trajectory; Generating a trajectory topological signature.

3. The illegal operation vehicle identification system based on the spatio-temporal graph neural network according to claim 1, characterized in that, The probability attention weighting layer is used for: Decomposing the trajectory into sub-trajectories of different time scales; Extracting features for each time scale to capture behavior patterns of different time granularities; Constructing a multi-scale feature matrix to represent the feature distributions of different time scales; Constructing a conditional probability model to quantify the importance of features for illegal operation judgment; Calculating the information gain of each feature as the basis of the attention weight; Designing three attention heads of time, space and spatio-temporal joint; Highlighting abnormal behavior features through the attention mechanism to improve the detection sensitivity.

4. The illegal operation vehicle identification system based on the spatio-temporal graph neural network according to claim 1, characterized in that, The spatio-temporal feature fusion layer is used for: Analyzing the complementarity and redundancy of features from different sources; Removing highly correlated features and retaining complementary information; Constructing a feature dependence graph to guide the feature fusion strategy; Adjusting the fusion weight according to urban characteristics, time periods and weather factors; Construct the combined features representing the starting area, travel path, and passing areas; Retain the high-order correlations among the features; Generate the final spatio-temporal combined feature vector.

5. The illegal operation vehicle identification system based on a spatio-temporal graph neural network according to claim 1, wherein The data preprocessing module is also used for: Normalize the time information to unify the time representation formats of data from different sources; Normalize the spatial information and map the location information to a unified geographic coordinate system; Normalize the vehicle information to unify the vehicle identification representations from different sources; Perform outlier detection and processing to identify and handle data that significantly deviates from the normal range; Complete the missing values and make reasonable inferences about the missing data based on spatio-temporal correlations; Verify the data consistency and check whether the same information from different sources is consistent.

6. The illegal operation vehicle identification system based on the spatio-temporal graph neural network according to claim 1, wherein The spatio-temporal graph construction module is also used for: Set the driving habits and travel time period attributes for the driver nodes; Set the geographical location and passing frequency attributes for the checkpoint nodes; Set the vehicle type and license plate color attributes for the vehicle nodes; Set the functional type and regional feature attributes for the location nodes; Construct the vehicle-checkpoint edges to represent the time and frequency of vehicles passing through the checkpoints; Construct the vehicle-location edges to represent the stay time and frequency of vehicles at the locations; Construct the driver-vehicle edges to represent the driving association intensity and time distribution; Construct the location-location edges to represent the travel frequency and passing time between locations; Implement dynamic graph updates to update the graph structure when new data arrives.

7. The illegal operation vehicle identification system based on spatio-temporal graph neural network according to claim 1, wherein The traffic feature extraction module is also used for: Calculate the travel frequency and count the number of trips within different time windows; Analyze the trip coherence and evaluate the spatio-temporal logical rationality between consecutive trips; Extract the parking patterns and identify the regular parking spots and duration distributions; Calculate the Gini coefficient of the trip trajectories to quantify the inequality of travel distances; Set the Gini coefficient threshold to determine the identity of professional drivers; Divide the urban area into several sectors; Construct the spatio-temporal feature sets using the boundary checkpoints of each sector; Cluster the boundary checkpoints into several clusters according to the feature similarity of the feature sets; Calculate the average stay time of checkpoints within the clusters and calculate the probability of illegal operation.

8. The illegal operation vehicle identification system based on spatio-temporal graph neural network according to claim 1, characterized in that, The system also includes: The passenger APP feature extraction module, connected to the data interaction module, is used to extract the user rating patterns, search behaviors, and payment records from the passenger APP operation behavior data and calculate the probability of illegal operation; The data fusion module is also used to fuse the probability of illegal operation.

9. The illegal operation vehicle identification system based on a spatio-temporal graph neural network according to claim 1, characterized in that The system also includes: The blockchain evidence storage module, connected to the intelligent decision-making module, is used to receive and encrypt and store the identification results of the illegal operation vehicles, including MAC addresses, usage times, license plates, and driving route information; Perform block storage according to a preset time period; Provide verifiable law enforcement evidence.

10. An illegal operation vehicle identification method based on a spatio-temporal graph neural network, using the system according to any one of claims 1-9, characterized in that, It includes: Connect to the mobile payment management platform, urban traffic control platform, and blockchain evidence storage platform through the data interaction module to obtain the checkpoint capture data, mobile payment platform trip data, urban traffic control platform data, and passenger APP operation behavior data; Unify the formats of the checkpoint capture data, the mobile payment platform trip data, the urban traffic control platform data, and the passenger APP operation behavior data through the data preprocessing module, and perform encoding standardization on each dimension of data; The spatio-temporal graph construction module constructs a spatio-temporal dynamic relationship graph containing driver nodes, checkpoint nodes, vehicle nodes, vehicle type nodes, trip nodes, and parking point nodes based on the encoded and standardized data; The topological enhanced hierarchical spatio-temporal attention network extracts spatio-temporal joint feature vectors from the spatio-temporal dynamic relationship graph. The topological enhanced hierarchical spatio-temporal attention network includes a topological space mapping layer, a probability attention weighting layer, and a spatio-temporal feature fusion layer; The traffic feature extraction module extracts travel frequency and trip coherence features based on the time-space trajectory of the checkpoint capture data, and calculates the probability of illegal operation; The data fusion module fuses the spatio-temporal joint feature vectors, the travel frequency, and the trip coherence features to extract candidate illegal operation vehicles and form an illegal operation vehicle library; The intelligent decision-making module calculates the weight coefficients of each dimension feature, and screens the vehicles in the illegal operation vehicle library based on the entropy value method to determine the illegal operation vehicles.

Images