Transport vehicle leaking stoppage and repayment big data analysis system and method

Abstract

The invention discloses a big data analysis system and method for leaking stoppage and repayment of transport vehicles. The big data analysis system comprises a traffic behavior monitoring module which collects vehicle tracks and load records in a road network traffic data flow, analyzes a traffic time distribution rule, calculates a path deviation index, compares the difference between actual load and declaration data, and generates an abnormal traffic index; and the fee evasion risk prediction module is used for positioning passing nodes of high-risk vehicles based on the abnormal passing indexes, analyzing a matching relationship between node characteristics and historical fee evasion events, evaluating a road network risk level and generating a fee evasion probability prediction result. According to the big data analysis system and method for leaking stoppage and repayment of the transport vehicle, a quantum key distribution and lightweight encryption fusion mechanism is constructed, full-link dynamic encryption of data transmission is realized, man-in-the-middle attack and data tampering are effectively resisted, the security of core indexes in links of acquisition, transmission and analysis is guaranteed, and a trusted data basis is provided for intelligent inspection.

Description

Technical Field

[0001] The present invention relates to the technical field of the intersection of smart transportation and data security, and specifically to a big data analysis system and method for leak plugging and reimbursement of transportation vehicles. Background Art

[0002] With the rapid development of intelligent transportation systems, highway operators are facing severe challenges from increasingly intelligent and covert toll evasion methods used by transport vehicles. Traditional inspection methods rely on manual verification and static data analysis, resulting in delayed responses and low recognition rates for toll evasion patterns. This is especially true in scenarios such as overloaded transport and interprovincial transport, where multi-source heterogeneous data, such as vehicle trajectory data and weighing information, is generated at high speed and high concurrency. This places higher demands on secure transmission and real-time processing from the data acquisition end to the analysis end.

[0003] Although the big data audit system currently widely used in the industry can achieve basic behavioral analysis, it has significant flaws in high-speed data security: First, sensitive data such as real-time vehicle trajectory and load fluctuations are easily subject to man-in-the-middle attacks during the interaction between the roadside unit (RSU) and the cloud, resulting in the tampering or forgery of key characteristic parameters, directly affecting the accuracy of the toll evasion risk model; second, the lack of fine-grained access control in the distributed storage of massive traffic records poses the risk of unauthorized retrieval of core indicators such as path deviation index and load fluctuation coefficient, which may provide a data basis for criminals to reverse-engineer the audit logic. More seriously, because some systems have not established dynamic encryption channels, high-risk vehicle information is easily intercepted in the transmission link during cross-regional data collaborative warning, rendering the dynamic interception strategy ineffective;

[0004] In addition, there is a contradiction between the security protection and in-depth analysis of real-time data streams in existing technologies: excessive encryption will reduce the efficiency of generating the traffic behavior feature matrix, while lightweight security strategies are difficult to resist replay attacks on the data streams of weighing equipment (such as forging load fluctuation signals to cover up toll evasion). This imbalance between security and real-time performance has become a key bottleneck restricting the improvement of the efficiency of smart road network leak plugging and reimbursement. Therefore, building a leak plugging and reimbursement system that takes into account both high-speed data security protection and intelligent analysis capabilities has become an urgent need to promote the modernization of transportation governance. Summary of the Invention

[0005] The purpose of the present invention is to provide a big data analysis system and method for transport vehicle leak plugging and supplementary payment, so as to solve the problem proposed in the above background technology that the system has not established a dynamic encryption channel, and high-risk vehicle information is easily intercepted in the transmission link during cross-regional data collaborative warning, resulting in the failure of the dynamic interception strategy.

[0006] To achieve the above objectives, the present invention provides the following technical solutions: a transport vehicle leak plugging and supplementary payment big data analysis system and method, comprising:

[0007] Perception layer: Integrates millimeter-wave radar (accuracy ±5cm), a dynamic weighing platform (sampling rate 100Hz), and a Beidou high-precision positioning module (positioning error <30cm) to achieve multimodal data collection of vehicle 3D profiles, real-time load, and trajectory coordinates;

[0008] Edge computing layer: A lightweight analysis model is deployed on the roadside edge server to perform primary feature extraction such as load fluctuation calculation and path deviation detection, with response time controlled within 500ms.

[0009] Secure transmission layer: Quantum key distribution (QKD) technology is used to establish a dynamic encryption channel, and the encryption key is updated every 30 seconds to ensure that the information security requirements of the ISO21434 standard are met during data transmission;

[0010] Cloud analysis layer: Build a distributed evasion risk prediction engine, integrating three analysis modules: XGBoost classifier (92.7% accuracy), time series pattern mining algorithm (LSTM network), and spatial propagation model (improved PageRank algorithm);

[0011] Application decision-making layer: Develop an intelligent inspection and dispatching system to support core functions such as automatic drone path planning (A* algorithm optimization), dynamic deployment of mobile checkpoints (game theory model), and automatic triggering of credit punishment (smart contracts).

[0012] To achieve the above objectives, the present invention provides the following technical solutions and a description of the process of transforming the evasion pattern matching algorithm:

[0013] Input layer transformation: original financial feature T i (Transaction amount / Frequency) → Traffic characteristics V i (nighttime traffic frequency / load fluctuation rate);

[0014] Risk event database Ri migration: enterprise financial violation records → historical fee evasion case feature database;

[0015] Increase the weight of the space-time dimension: introduce the time decay factor λ = 0.85, and the formula is modified as follows:

[0016]

[0017] Dynamic threshold adjustment: When the road network traffic volume is greater than 500 vehicles / hour, the matching threshold is automatically lowered by 20%;

[0018] Output layer expansion: Added risk path topology map generation function to visualize high-risk road sections (GIS heat map rendering);

[0019] Classification label library of fee evasion methods: overstay (35%), false declaration (42%), equipment interference (23%).

[0020] To achieve the above objectives, the present invention provides the following technical solutions, including a traffic behavior monitoring module:

[0021] This module overcomes the technical problem of low reliability of weighing data in high-noise environments and proposes a three-level data cleaning mechanism:

[0022] 1. Signal denoising: Wavelet packet transform is used to decompose the original weighing signal, and the sym4 wavelet basis function is selected for 6-layer decomposition. The signal-to-noise ratio of the reconstructed signal is improved to 35dB;

[0023] 2. Outlier elimination: A dynamic threshold model based on Mahalanobis distance is established. When the fluctuation range of three consecutive sampling points exceeds μ±3σ, the data verification process is triggered.

[0024] 3. Space-time alignment: A timestamp compensation algorithm is designed to solve the millisecond-level time synchronization problem between GPS positioning data and weighing records, and the alignment error is controlled within ±50ms.

[0025] To achieve the above objectives, the present invention provides the following technical solutions, including a fee evasion risk prediction module:

[0026] This module builds a risk assessment system that integrates spatiotemporal features:

[0027] 1. Time decay factor: An exponential decay function is introduced to dynamically weight historical fare evasion events, ensuring that the model can capture long-term behavioral patterns while quickly responding to new fare evasion methods.

[0028] 2. Spatial Propagation Model: Improves the traditional PageRank algorithm, increases the risk transmission weight factor ω, and accurately identifies key nodes in cross-regional fee evasion paths (improves identification accuracy by 28%).

[0029] 3. Ensemble learning framework: We built a stacking model using random forest (100 decision trees) as the base learner and logistic regression as the meta-learner, achieving a predictive performance of F1-score 0.89 on a provincial road network test set.

[0030] To achieve the above objectives, the present invention provides the following technical solutions, including a dynamic interception strategy module:

[0031] This module proposes an audit resource optimization model based on game theory:

[0032] 1. Multi-objective planning: Establish a joint objective function of minimizing response time and maximizing audit efficiency, and use the NSGA-II algorithm to solve the Pareto optimal solution set.

[0033] 2. Dynamic deployment algorithm: Design an improved ant colony algorithm to achieve second-level updates of inspection equipment deployment plans in a 500-node road network (solution time < 3s);

[0034] 3. Virtual-reality linkage mechanism: Develop a digital twin simulation platform that can predict the possible evasive paths taken by toll-evading vehicles 30 minutes in advance, guiding the preventive deployment of physical checkpoints.

[0035] To achieve the above objectives, the present invention provides the following technical solutions, including a path deviation calculation algorithm:

[0036] A quantitative evaluation model integrating actual mileage and declared routes is proposed:

[0037]

[0038] in:

[0039] The actual mileage of the i-th road segment;

[0040] The reported mileage of the i-th road section;

[0041] σ u : speed standard deviation;

[0042] u: average speed.

[0043] By introducing a speed fluctuation factor, the algorithm can effectively distinguish between reasonable detours and intentional toll evasion (AUC value reaches 0.93).

[0044] To achieve the above objectives, the present invention provides the following technical solutions, including a risk path propagation algorithm:

[0045] Improved PageRank algorithm formula:

[0046]

[0047] New parameters:

[0048] ω qp : Risk transmission weight from node q to p (dynamically adjusted between 0.8-1.2);

[0049] C(q): out-degree correction factor of node q.

[0050] The model successfully identified 83% of toll evasion hotspots in provincial border areas, an improvement of 37 percentage points compared to traditional methods.

[0051] To achieve the above objectives, the present invention provides the following technical solutions, including an audit resource scheduling algorithm:

[0052] Establish a multi-agent game model:

[0053]

[0054] Constraints:

[0055] (total resource constraints);

[0056] (Coverage constraint).

[0057] This model reduces inspection costs in actual road network measurements and increases the interception rate of toll evasion.

[0058] Dynamic encryption transmission mechanism: The first fusion solution of quantum key distribution and lightweight national encryption algorithm, while ensuring a transmission rate of 200Mbps, achieving ciphertext anti-quantum cracking strength >128 bits;

[0059] Spatiotemporal joint analysis model: Constructing a dual-dimensional evaluation system of "time decay factor + spatial transmission weight" to overcome the challenge of identifying cross-regional fee evasion;

[0060] Game optimization scheduling system: establishes an asymmetric game model between the auditor and the toll evader to achieve the optimal dynamic configuration of road network resources;

[0061] Credit Punishment Smart Contract: Develop an automatically executed contract based on blockchain, reducing the processing time for overdue payment cases from 72 hours to 4 hours.

[0062] Compared with the prior art, the transport vehicle leak plugging and supplementary payment big data analysis system and method of the present invention have the following beneficial effects:

[0063] Construct a fusion mechanism of quantum key distribution and lightweight encryption to achieve dynamic encryption of the entire data transmission link, effectively resist middleman attacks and data tampering, ensure the security of core indicators in the collection, transmission and analysis of each link, and provide a trusted data foundation for intelligent inspection. Through the joint spatiotemporal analysis model, it accurately captures abnormal patterns in vehicle traffic behavior, combines improved algorithms to achieve intelligent deduction of cross-regional evasion paths, significantly improves the recognition accuracy of new evasion methods, and provides forward-looking decision support for dynamic deployment. Based on asymmetric game models and digital twin technology, it realizes real-time optimization of inspection equipment deployment plans, breaking through the time constraints of traditional fixed checkpoints. It can overcome the limitations of empty space, form a multi-dimensional inspection network while reducing manpower input, greatly improve the efficiency of road network resource utilization, open up the complete chain from behavior monitoring, risk assessment to credit punishment, automatically trigger the hierarchical response mechanism through smart contracts, strengthen the cross-departmental collaborative disposal capabilities, form a monitoring and early warning-precise interception-credit punishment governance closed loop, effectively curb the recurrence of fare evasion, build a dynamic iterative model of driver group behavior portrait, deeply explore the evolution of fare evasion methods, continuously optimize the feature library and algorithm parameters, ensure that the system has the adaptive ability to deal with new fare evasion modes, and promote the intelligent upgrade of the transportation governance system. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 This is a flow chart of the overall architecture of the system of the present invention;

[0065] Figure 2 This is a flow chart of traffic behavior monitoring in the present invention;

[0066] Figure 3 This is a flow chart of the risk assessment of fee evasion according to the present invention;

[0067] Figure 4 This is a flow chart of the dynamic interception execution of the present invention;

[0068] Figure 5 This is a flowchart of the credit punishment execution of the present invention. DETAILED DESCRIPTION

[0069] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the implementation regulations described are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0070] The system consists of five core modules:

[0071] 1. Traffic behavior monitoring module: collects vehicle trajectory and load data in real time, analyzes traffic patterns and generates abnormal indicators;

[0072] 2. Toll evasion risk prediction module: This module uses machine learning models to assess the probability of toll evasion and locate high-risk road sections.

[0073] 3. Dynamic interception strategy module: optimizes the deployment of inspection equipment and realizes intelligent allocation of road network resources;

[0074] 4. Behavior pattern analysis module: Build a profile database of driver fare evasion behaviors;

[0075] 5. Repayment response execution module: automatically triggers the hierarchical repayment process and connects to the credit punishment platform.

[0076] Example 1

[0077] A logistics company declared to transport a batch of mechanical equipment (declared load capacity 40 tons), but during the actual transportation, the route was frequently switched and the weighing equipment was interfered with.

[0078] 1. Data collection: The weighing equipment uploads the load data stream in real time (e.g. 38.2 tons → 15.6 tons → 42.3 tons);

[0079] The GPS track showed that the vehicle detoured onto the undeclared County Road X203.

[0080] 2. Abnormal indicator calculation:

[0081] a. Load fluctuation coefficient:

[0082]

[0083] Trigger an exception flag.

[0084] b. Path deviation:

[0085]

[0086] Determined to be an abnormal detour.

[0087] 3. The output result is that the system generates an abnormal traffic report, marks the vehicle as a high-risk vehicle, and synchronizes it to the inspection terminal.

[0088] The technical effect is that the accuracy of load fluctuation detection is increased to 92%, and the response time for path abnormality judgment is shortened to 5 seconds.

[0089] Example 2

[0090] A toll evasion gang frequently used the intersection of Provincial Road S301 and Expressway G7 to switch routes and avoid charging.

[0091] 1. Feature matching:

[0092] The characteristics of historical toll evasion cases were extracted: nighttime travel (22:00-4:00), load fluctuation coefficient > 0.4, and path deviation > 0.2.

[0093] 2. Risk path analysis:

[0094] Use the improved PageRank algorithm to calculate the node influence value:

[0095]

[0096] The attenuation factor d = 0.85, identifying the S301-G7 intersection as the core risk node (PR value = 1.8).

[0097] 3. Collaborative early warning:

[0098] When the PR values of the three adjacent nodes are all greater than 1.5, the system automatically triggers a cross-regional joint inspection command, dispatches drones and mobile checkpoints to block the area, increases the high-risk path identification coverage by 40%, and improves the cross-regional inspection response efficiency by 60%.

[0099] Example 3

[0100] A certain fleet took advantage of the blind spot of nighttime inspections and frequently evaded tolls between 22:00 and 4:00.

[0101] 1. Time distribution modeling:

[0102] a. Count the historical vehicle travel time series {t_1, t_2, ..., t_n} and construct a time distribution histogram;

[0103] b. Calculate the proportion of nighttime traffic frequency:

[0104]

[0105] When N_nightda is greater than the threshold, it is considered abnormal.

[0106] 2. Risk prediction:

[0107] Introducing a time decay factor λ = 0.85, dynamically weighting historical evasion events:

[0108]

[0109] Output risk level labels to mark high-risk vehicles.

[0110] 3. Resource Scheduling:

[0111] Generate inspection intensity parameters based on real-time traffic flow F and risk value R:

[0112] Q=F·R

[0113] Dynamically adjust the frequency of drone patrols and checkpoint duration.

[0114] Example 4

[0115] The driver interferes with the weighing equipment by braking suddenly, pressing the multi-axis scale, etc., resulting in abnormal load data. 1. Data cleaning and reconstruction:

[0116] Perform wavelet transform to denoise the original weighing signal:

[0117]

[0118] The average load value of document pickup is extracted as valid data.

[0119] 2. Behavioral pattern analysis:

[0120] Constructing a "device interference" feature profile:

[0121] The variance of the loading sequence is significantly higher than normal;

[0122] Path deviation is strongly correlated with speed fluctuation (Pearson coefficient > 0.7).

[0123] Example 5

[0124] A certain enterprise has evaded fees many times and refused to make up the difference, and cross-departmental joint punishment needs to be initiated.

[0125] 1. Credit data association:

[0126] Automatically match the unified social credit code of the enterprise and retrieve multi-source data such as industry and commerce, taxation, etc.

[0127] 2. Hierarchical response mechanism:

[0128] Generate a payment notice and calculate late payment fees:

[0129] P=P0·(1+r) t

[0130] Where r is the daily late payment fee, and t is the number of days overdue;

[0131] For enterprises that are overdue for more than 30 days, they will be automatically pushed to the "Credit China" platform for joint punishment.

[0132] 3. Effect evaluation:

[0133] Calculate indicators such as the repayment rate and punishment response cycle, and iteratively optimize the credit assessment model.

[0134] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0135] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is limited by the appended claims and their equivalents.

Claims

1. A big data analysis system and method for leak plugging and supplementary payment of transport vehicles, characterized in that: The system comprises: Traffic behavior monitoring module: collects vehicle trajectories and load records from the road network traffic data stream, analyzes the distribution of travel time, calculates the path deviation index, compares the actual load with the reported data, and generates abnormal traffic indicators; Toll evasion risk prediction module: Based on the abnormal traffic indicators, it locates the traffic nodes of high-risk vehicles, analyzes the matching relationship between node characteristics and historical toll evasion events, evaluates the road network risk level, and generates toll evasion probability prediction results; Dynamic interception strategy module: Based on risk prediction results, it identifies road network nodes that require key monitoring, dynamically deploys mobile inspection equipment, and generates optimized road network resource allocation plans; Behavior pattern analysis module: Analyzes the travel time preferences and route selection patterns of driver groups, and builds a profile database of toll evasion behavior characteristics; Supplementary payment response execution module: automatically links to the corporate credit database, generates graded supplementary payment notices and synchronizes them to the road administration law enforcement platform.

2. A transport vehicle leak plugging and supplementary payment big data analysis system and method according to claim 1, characterized in that: Abnormal traffic indicators include nighttime traffic frequency, load fluctuation coefficient, and path deviation; The evasion probability prediction results include a risk path topology map and evasion method classification labels; The road network resource optimization configuration plan specifies the drone patrol routes and the linkage strategy of fixed checkpoints.

3. The big data analysis system and method for transport vehicle leak plugging and supplementary payment according to claim 1 is characterized by: The traffic behavior monitoring module includes: Load fluctuation analysis submodule: real-time analysis of weighing equipment data stream, calculation of the load change rate of continuous pass records, and triggering an abnormal flag when the fluctuation coefficient is greater than 0.3; Path feature extraction submodule: Generate path deviation index based on GIS trajectory data (L a is the actual mileage, L e (for reporting mileage), D P When it is greater than 0.15, it is determined as an abnormal path; Declaration data verification submodule: Call the provincial large-scale transportation permit database to compare the actual passage time window with the declaration plan.

4. A transport vehicle leak plugging and supplementary payment big data analysis system and method according to claim 1, characterized in that: The fee evasion risk prediction module includes: Multi-dimensional feature matching submodule: Vehicle traffic feature vector [D P ,M,θ] inputs the trained fee evasion classification model and outputs the risk level label, where M is the load fluctuation coefficient and θ is the time decay factor definition; Path propagation analysis submodule: Calculates the risk path influence value based on an improved algorithm Dynamically identify the core sections that require interception; When the θ of three adjacent road network nodes is greater than 1.5, the collaborative warning generation submodule automatically initiates a cross-regional joint inspection.

5. The big data analysis system and method for transport vehicle leak plugging and supplementary payment according to claim 1 is characterized by: The dynamic interception strategy module includes: Mobile device scheduling algorithm: Optimize the UAV deployment position for the objective function, where t resp is the response delay, cost is the scheduling cost; Checkpoint dynamic start-stop submodule: Dynamically adjusts the checkpoint inspection intensity based on the product of real-time traffic flow and risk value Q = F·R (F is the traffic flow, R is the risk value).

6. A method for analyzing the leakage and supplementary payment of large-scale transport vehicles, characterized in that: Including steps: S1: Integrate toll collection system, surveillance video, and GPS trajectory data to construct a traffic behavior feature matrix; S2: Identify high-risk vehicles through toll evasion pattern matching algorithms and generate a risk path heat map; S3: Calculate the optimal audit resource allocation plan based on the dynamic programming model; S4: Analyze the evolution of driver behavior patterns and iteratively update the feature profile library; S5: Connect to the Credit China platform and automatically initiate a joint punishment process for companies that refuse to make up the payments.

Images