Driving training hour authenticity monitoring method and system based on GPS and base station double verification

Abstract

The application provides a driving training hour authenticity monitoring method and system based on GPS and base station double verification, through vehicle-mounted equipment integrated collection of GPS, base station, vehicle operation and other data, multi-dimensional, deep cross verification and abnormal mode recognition are carried out in the cloud, so that various cheating behaviors are effectively prevented, and the authenticity and effectiveness of the training hour record are ensured, GPS, base station and vehicle data are collected through the vehicle-mounted equipment, cross verification is carried out in the cloud: the matching degree of the GPS position and the base station coverage range is calculated, the logic contradiction between the motion / static state and the base station switching is detected, and the position is independently compared by using multi-base station triangulation; then, the displacement total amount and the base station signal fluctuation are used for secondary confirmation of static cheating; finally, the GPS trajectory, the base station consistency and the vehicle data are weighted and scored, and if the score is lower than the threshold value, an alarm is given and the abnormality is recorded.

Description

Technical Field

[0001] This invention relates to the field of driver training monitoring, and more specifically, to a method and system for monitoring the authenticity of driver training hours based on dual verification of GPS and base stations. Background Technology

[0002] In the driver training industry, trainees must complete a prescribed number of training hours before taking the exam, and the authenticity of these hours directly impacts training quality and road safety. Current monitoring primarily relies on timing terminals installed in training vehicles, using GPS positioning to determine if the vehicle is on the training ground and moving. However, various cheating methods have emerged, such as using GPS signal simulators to falsify driving trajectories, using odometers to artificially inflate mileage, or placing timing devices in fixed positions to generate false training hours.

[0003] To address the aforementioned issues, existing solutions attempt to incorporate multi-source data to enhance anti-counterfeiting capabilities. For example, patent application CN202011031766.8 discloses a positioning method and system for driver training systems. This method requires trainees to carry a mobile terminal, which connects to the driver training terminal via Bluetooth and performs dynamic key authentication. The driver training terminal then obtains its own and the mobile phone's location data, which may include GPS, base station data, etc., and uses AI algorithms to select or fuse reliable location information. This solution increases the difficulty of cheating to some extent, but still has significant limitations: First, it relies on students' mobile phones, which increases the burden on students, and issues such as the stability and compatibility of mobile phone Bluetooth connections can affect the reliability of the system. Secondly, its essence is still to select or simply merge the two positioning source data, lacking in-depth verification of the inherent logical consistency between GPS positioning behavior and base station cellular network behavior, and unable to effectively identify advanced cheating methods such as "GPS trajectory seems normal, but the base station switching mode seriously contradicts the actual movement pattern". Third, the solution does not involve the fusion of vehicle operating data such as OBD vehicle speed, engine status and motion sensor data, resulting in a single judgment dimension and a high risk of misjudgment and omission. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for monitoring the authenticity of driver training hours based on dual verification of GPS and base stations. By integrating data from GPS, base stations, and vehicle operation through in-vehicle equipment, multi-dimensional and in-depth cross-verification and abnormal pattern recognition are performed in the cloud, thereby effectively preventing various cheating behaviors and ensuring the authenticity and validity of training hour records.

[0005] The embodiments of the present invention are implemented as follows: A method for monitoring the authenticity of driver training hours based on dual verification of GPS and base stations includes the following steps: Data collection steps: During student training, GPS positioning data, base station information data, and vehicle operation data are continuously collected using onboard training equipment installed in the training vehicle. Base station information data includes at least the cell ID, location area code, signal strength, and a list of neighboring base stations and their signal strengths. The GPS positioning data and base station information data are collected every 10-60 seconds, while the vehicle operation data is collected every 1-10 seconds. The dual verification process involves cross-verifying GPS positioning data and base station information data. This cross-verification includes: calculating the matching degree between the GPS positioning location and the theoretical coverage area of ​​the currently connected base station; and determining data anomalies if the GPS positioning data shows vehicle movement while the base station information data indicates no switching, or if the GPS positioning data shows vehicle stationary while the base station information data indicates frequent switching; and performing triangulation calculations using the signal strength of multiple base stations in the adjacent base station list to obtain a location estimate independent of GPS positioning, which is then compared and verified with the GPS positioning location. Location Anomaly Detection Steps: Calculate the total displacement obtained from GPS positioning data within a training period. If the total displacement is lower than a preset displacement threshold and vehicle operation data indicates no effective mileage, then a secondary confirmation is performed using base station information data. If base station information data indicates that the base station does not perform any handovers and the signal strength fluctuation is lower than a preset fluctuation threshold during the training period, then the vehicle-mounted training equipment is determined to be in a fixed position, indicating suspected cheating. In the location anomaly detection steps, the preset displacement threshold is 500 meters. The secondary confirmation specifically involves: if the base station does not perform any cell handovers during the training period and the change in signal strength is always less than a preset decibel value, then the equipment is determined to be in a fixed position. Multi-dimensional fusion scoring steps: Calculate the comprehensive credibility score S based on the preset multi-dimensional verification index system. The index system includes at least the GPS trajectory rationality score, the GPS and base station data consistency score, and the vehicle operation data authenticity score; and obtain the comprehensive credibility score S by weighted summation of each score. Anomaly handling steps: When the overall credibility score S is lower than the preset credibility threshold, the corresponding learning hour record is determined to be abnormal, the abnormal information and abnormal evidence are recorded, and displayed through the supervision platform for manual review.

[0006] In a preferred embodiment of the present invention, before the dual verification step, a data preprocessing step is also included: synchronizing and aligning the collected GPS positioning data, base station information data and vehicle operation data according to timestamps, and filtering the GPS trajectory data to remove abnormal positioning points.

[0007] In a preferred embodiment of the present invention, the dual verification step further includes GPS trajectory rationality verification, which includes: analyzing whether the GPS trajectory is within the preset training site range; calculating the distance and speed between adjacent GPS positioning points, determining whether the speed change is continuous and whether the acceleration is within a reasonable range; and detecting whether the GPS trajectory has abnormal jumps, instantaneous shifts or repetitions.

[0008] In a preferred embodiment of the present invention, in the multi-dimensional fusion scoring step, the GPS trajectory rationality score is recorded as A, the GPS and base station data consistency score is recorded as B, and the vehicle operation data authenticity score is recorded as C; the formula for calculating the comprehensive credibility score S is: S = w1× A + w2× B + w3× C; Where w1, w2, and w3 are weight coefficients, satisfying w1 + w2 + w3 = 1, and the range of each weight's proportion is as follows: w1∈ [0.3, 0.4], w2∈ [0.4, 0.5], w3∈ [0.2, 0.3].

[0009] In a preferred embodiment of the present invention, the method further includes: deploying a lightweight verification algorithm on an on-board training device to perform preliminary anomaly screening on the collected data, and uploading relevant detailed data to a cloud verification server for in-depth analysis only when suspicious situations are detected by the preliminary anomaly screening.

[0010] In a preferred embodiment of the present invention, the vehicle operation data also includes vehicle motion pattern data collected by accelerometers and gyroscopes integrated in the vehicle training equipment; in the multi-dimensional fusion scoring step, a score is also calculated based on the consistency between the vehicle motion pattern data and the GPS positioning data, in order to participate in the calculation of the comprehensive credibility score.

[0011] In a preferred embodiment of the present invention, the above method further includes a continuous biometric authentication step: the student is initially authenticated at the start of training, and during the training process, multiple identity verifications are randomly initiated through the biometric recognition module integrated in the vehicle-mounted training equipment, and the verification results are used to assist in judging the authenticity of the training hour records.

[0012] This invention also provides a driver training hour authenticity monitoring system based on GPS and base station dual verification, which includes: The vehicle-mounted training equipment, installed on the training vehicle, integrates a GPS positioning module and a mobile communication module. It is used to collect GPS positioning data, base station information data, and vehicle operation data during the training period and upload them through the mobile communication network. Among them, the base station information data includes at least the cell ID, location area code, signal strength, and a list of neighboring base stations and their signal strengths of the currently connected base station. A cloud-based verification server, connected to the network of the vehicle-mounted training equipment, is used to receive data and execute the method described in any of the preceding descriptions; The monitoring platform is connected to a cloud-based verification server to receive and display study hour records, anomaly alerts, and statistical data.

[0013] In a preferred embodiment of the present invention, the vehicle-mounted training device also integrates an accelerometer and a gyroscope for collecting vehicle motion pattern data; when executing the method, the cloud verification server also uses the vehicle motion pattern data for fusion verification.

[0014] In a preferred embodiment of the present invention, the vehicle-mounted training device also integrates a biometric identification module for randomly verifying the identity of trainees multiple times during the training process; when executing the method, the cloud verification server also combines the identity verification results to determine the validity of the training hours.

[0015] The beneficial effects of the embodiments of the present invention are: 1. The system is simpler and more reliable because it does not rely on students' mobile phones. All data acquisition functions are integrated into the vehicle-mounted training equipment, which avoids compatibility and connection stability issues of students' devices and reduces the complexity of operation and maintenance. This is fundamentally different from the solution in CN202011031766.8 that relies on mobile terminals.

[0016] 2. A significant improvement in anti-cheating capabilities. This invention does not simply obtain base station location values, but rather delves into the inherent physical relationship between GPS trajectory behavior and base station cellular network handover behavior: movement is always accompanied by base station handover, and base stations should remain stable when stationary. Through this logical consistency check, even if a cheater uses a GPS simulator to perfectly forge a driving trajectory, they cannot synchronously simulate a base station handover pattern that conforms to real movement patterns, significantly raising the threshold for cheating; supplemented by independent position estimation using multi-base station triangulation, a third layer of verification is formed.

[0017] 3. Higher accuracy through multi-dimensional data fusion. This invention integrates multiple indicators, including GPS trajectory rationality, GPS-base station behavior consistency, vehicle OBD data, motion sensor data, and continuous biometric authentication, to construct a scientifically weighted comprehensive credibility scoring system. This effectively reduces the vulnerability of being deceived by a single data source and lowers the rates of misjudgment and missed judgment.

[0018] 4. A dedicated detection logic was designed for the fixed-position fake learning time scenario, with multiple conditions for progressive confirmation, to accurately identify cheating behavior when the device does not move with the vehicle. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the system composition of the present invention.

[0021] Figure 2 This is a schematic diagram of the main flow of the method of the present invention.

[0022] Figure 3 This is a logical diagram illustrating the cross-validation of GPS and base station data and the enhanced verification of triangulation.

[0023] Icons: 001 - Vehicle-mounted training equipment; 002 - Cloud verification server; 003 - Monitoring platform. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0025] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0026] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0027] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided to make this application more comprehensive and complete, and to fully convey the concept of the exemplary embodiments to those skilled in the art.

[0028] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0029] The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily have to be performed in the described order. For example, some operations / steps can be broken down, while others can be combined or partially combined; therefore, the actual execution order may change depending on the specific circumstances.

[0030] First Embodiment Please refer to Figure 1-3 This embodiment provides a driver training hour authenticity monitoring system based on GPS and base station dual verification, including vehicle-mounted training equipment 001, cloud verification server 002 and monitoring platform 003.

[0031] The vehicle-mounted training device 001 is installed in a training vehicle and integrates a GPS positioning module, a 4G / 5G mobile communication module, an OBD data acquisition interface, an accelerometer, a gyroscope, and a camera (for facial recognition). The device acquires real-time vehicle speed, engine RPM, engine status, and cumulative mileage through the OBD interface; it acquires location data such as latitude and longitude, speed, direction angle, and timestamp through the GPS module; and it obtains information about the current serving base station and neighboring base stations through the baseband processor of the mobile communication module, including Cell ID, LAC, RSSI, a list of neighboring base stations, and their signal strength. The Cell ID is the cell ID of the currently connected base station, the LAC is the location area code, and the RSSI is the signal strength. Simultaneously, it collects three-axis acceleration and angular velocity data through the built-in accelerometer and gyroscope. The vehicle-mounted training device 001 connects to the cloud verification server 002 via a mobile communication network or WiFi.

[0032] The cloud-based verification server 002 is responsible for receiving and storing all uploaded data and running real-time verification algorithms. The monitoring platform 003 connects to the cloud server via the network to provide driving school administrators and regulatory departments with visualized learning hour records, anomaly alerts, and data statistics query functions.

[0033] Reference Figure 2 The main process, as described in this embodiment, is implemented according to the following steps: Step 1: Data Collection Before training, trainees authenticate their identity using the facial recognition camera on the in-vehicle device. Once authentication is successful, the training time recording function is activated. During training, the device collects various types of data at set frequencies: GPS data and base station information are collected every 10-60 seconds (15 seconds in this embodiment), vehicle operation data (such as vehicle speed, engine status, etc.) are collected every 1-10 seconds (2 seconds in this embodiment), and accelerometer / gyroscope data are collected every 1 second. All data is packaged and uploaded to the cloud in real time after being accompanied by a unified timestamp and the trainee's ID and vehicle ID.

[0034] Taking a training session as an example, an instance of raw data collected in one session is as follows: (a) GPS positioning data: timestamp 14:30:15, longitude 104.06579, latitude 30.57215, speed 22km / h, orientation angle 85°, positioning accuracy 3m; (b) Base station information data: timestamp 14:30:15, serving base station Cell ID=20471, LAC=14350, RSSI=-68dBm; neighboring base station list includes: Cell ID=20472, RSSI=-75dBm, Cell ID=20473, RSSI=-80dBm, Cell ID=20115, RSSI=-85dBm, etc. (c) Vehicle operation data: timestamp 14:30:15, vehicle speed 21km / h, engine speed 1800rpm, engine status = started.

[0035] Step 2: Data Preprocessing and Synchronization After receiving the data, the cloud server first synchronizes and aligns the GPS track points, base station information records, vehicle speed, and other data according to the timestamps, forming a complete data sequence indexed by the time axis. For example, the GPS data, base station data, and vehicle data at 14:30:15 are associated as the same time-aligned record. Next, Kalman filtering is applied to the GPS track to remove isolated points with severe drift or sudden drops in positioning accuracy. For example, if the accuracy of a point suddenly jumps from 3m to 50m, it is filtered out as a noise point.

[0036] Step 3: Verification of GPS Track Reasonableness This verification serves as the basis for scoring indicator A. Specifically, it includes: (1) Electronic fence verification: Determine whether each GPS point falls within the preset polygonal area of ​​the driving school training ground. If the sampling points in the ground account for more than 95% of the total training period, the score for this item is 1.0; if the percentage is between 80% and 95%, the score is 0.6-0.9; if it is below 80%, the score is below 0.5.

[0037] (2) Kinematic characteristic verification: Calculate the distance Δd and time difference Δt between two adjacent GPS points to obtain the velocity v = Δd / Δt between the points. Check whether the velocity is within the normal range of 0-80km / h and whether the acceleration a = Δv / Δt is less than the preset rapid acceleration / deceleration threshold. In this embodiment, the threshold is 5m / s². Statistically count the proportion of sampling points outside the range. The lower the proportion, the higher the score. At the same time, check whether there are jump points in the trajectory with a single displacement of more than 300 meters, or repeated points with more than 5 consecutive sampling periods (i.e., more than 75 seconds) at the same coordinate; if so, deduct points for this item. Based on the above items, the GPS trajectory rationality score A is obtained, with a full score of 1.

[0038] Step 4: Dual verification of GPS and base station data and enhancement of triangulation positioning This step is the core innovation, and its result forms the main basis for a score of B. (Refer to...) Figure 3 The logic is illustrated below. This step includes three sub-verification processes, which will be described in detail below with algorithm details and specific examples.

[0039] Sub-step 4.1: Calculation of GPS and base station coverage matching degree The cloud-based verification server 002 pre-stores a database of base station geographic locations, recording the latitude and longitude coordinates and typical coverage radius (700 meters in urban areas and 1500 meters in suburban areas) corresponding to each base station Cell ID. For a GPS location point at a given sampling time, the following judgment is performed: Retrieve the latitude and longitude coordinates of the serving base station from the database at that moment, and denote the base station coordinates as B. lon B lat GPS location is G lon G lat ; Calculate the spherical distance d between the GPS location point and the base station using the Haversine formula: a = sin²(Δlat / 2) + cos(lat1)×cos(lat2)×sin²(Δlon / 2); c = 2 × atan2(√a, √(1-a)); d = R × c; Where R is the Earth's radius, taken as 6371 km, lat1 and lat2 are the latitudes of the two points, and Δlat and Δlon are the difference in latitude and longitude between the two points, respectively.

[0040] The calculated distance d is compared with the coverage radius R of the base station. cell Comparison: If d ≤ R cell If d > R, then the sampling point is marked as "match"; cell If it does not match, it will be marked as "not a match".

[0041] Statistically analyze the matching results of all N sampling points throughout the entire training period and calculate the matching rate P. match = N match / N, and press P match The mapping is to sub-score B1. For example, P match ≥ 95% gets 1.0 point, 90%-95% gets 0.8 points, 80%-90% gets 0.6 points, and below 80% gets 0.4 points or less.

[0042] Example: At a certain data collection time, the GPS location is (104.06579, 30.57215), and the current serving base station coordinates are (104.06612, 30.57180). The distance between the two points is calculated to be 68 meters using the Haversine formula. The base station is located in an urban area, with a preset coverage radius of 700 meters. Since 68 meters ≤ 700 meters, this point is marked as a match. If a total of 40 points are collected during the entire learning period, and 38 of them are matches, then P... match =95%, B1=1.0 point.

[0043] Sub-step 4.2: Detection of GPS motion status and base station handover behavior logic This sub-step identifies logical contradictions between the motion state reflected by GPS speed and base station cell handover events by comparing the two.

[0044] First, the GPS velocity sequence and base station Cell ID sequence are segmented and labeled: Motion segment determination: When the speed of 3 consecutive GPS sampling points is >10km / h, this time window is marked as "motion segment"; Determination of stationary segment: When the speed of 3 consecutive GPS sampling points is less than 2 km / h, this time window is marked as a "stationary segment".

[0045] Then, analyze the base station handover behavior: Base station handover event definition: A change in the Cell ID between two consecutive base station information collections is recorded as a base station handover.

[0046] Anomaly detection rule one, no handover during movement: For each "movement segment", if the segment lasts for more than 5 minutes, covering at least 20 base station sampling points, and the number of base station handovers during this period is 0, it is judged as "no handover during movement anomaly".

[0047] Anomaly detection rule two, frequent handover during stationary periods: For each "stationary segment", if the duration of the segment exceeds 5 minutes, but the base station undergoes 3 or more handovers during that period, it is determined to be an "anomaly of frequent handover during stationary periods".

[0048] Finally, the number of abnormal segments and their cumulative duration throughout the entire training period are counted, and the sub-score B2 is calculated. Specifically, the percentage of abnormal duration R is defined. anomaly = Cumulative duration of abnormal segments / Total training duration. If R... anomaly =0 (no abnormalities), B2=1.0 point; R anomaly 0-10% gets 0.7-0.9 points; 10%-30% gets 0.4-0.6 points; over 30% gets 0.3 points or less.

[0049] Example: Trainee A's training record shows that from 14:30 to 14:50, a period of 20 minutes, the GPS speed fluctuated between 15-30 km / h, which the system determined to be a "movement segment." However, analysis of base station data shows that the serving base station Cell ID remained consistently 20471 during this period, with zero handovers. According to the detection rules, this is classified as an "abnormality of no handover during movement," and the entire 20-minute period is counted as abnormal time. If the total training time is 60 minutes, then R... anomaly =20 / 60≈33.3%, corresponding to B2=0.3 points.

[0050] Sub-step 4.3: Independent verification of multi-base station triangulation This sub-step utilizes the signal strength of multiple base stations in the adjacent base station list and their known geographical locations to estimate the vehicle's position using a triangulation algorithm, and then independently compares it with the GPS positioning coordinates. The specific algorithm flow is as follows: (1) Selection of reference base stations: For each sampling time, select the three base stations with the strongest signal strength (RSSI) from the list of neighboring base stations, and together with the current serving base station, obtain a total of 4 candidate base stations. Prioritize the selection of three base stations with known geographical locations and not collinear as the calculation base stations. Let the latitude and longitude coordinates of the three base stations be (BaseX1, BaseY1), (BaseX2, BaseY2), and (BaseX3, BaseY3), respectively.

[0051] (2) RSSI ranging conversion: Using the Okumura-Hata propagation model or the simplified logarithmic distance path loss model, the RSSI values ​​of each base station are converted into estimated distances from the terminal to the base station. The simplified model formula is: RSSI = -10 × n × log 10 (d) + A, Where d is the distance in meters, n is the path loss exponent (3.0-3.5 in urban environments), and A is typically a constant signal strength at 1 meter, calibrated based on base station antenna parameters, usually around -30 dBm. The distance estimate can then be derived from this. d = 10^((A - RSSI) / (10 × n)).

[0052] Example illustration: Select three base stations, with known coordinates and RSSI as follows: Base station 1: Coordinates (104.0650, 30.5710), RSSI = -68dBm; Base station 2: coordinates (104.0680, 30.5720), RSSI = -75dBm; Base station 3: coordinates (104.0630, 30.5740), RSSI=-80dBm.

[0053] Given parameters A = -30dBm and n = 3.2, then: d1 = 10^((-30 - (-68)) / (10×3.2)) = 10^(38 / 32) = 10^1.1875 ≈ 15.4 meters; d2 = 10^((-30 - (-75)) / 32) = 10^(45 / 32) = 10^1.40625 ≈ 25.5 meters; d3= 10^((-30 - (-80)) / 32) = 10^(50 / 32) = 10^1.5625 ≈ 36.5 meters.

[0054] (3) Triangulation solution: Based on the coordinates of the three base stations and the corresponding distance estimates d1, d2, and d3, the vehicle position estimate is solved using the least squares method. , The basic principle is as follows: Establish the observation equation system. For each base station i, we have: ( - BaseX i )² + ( - BaseY i )² = d i ²; Subtracting the three equations pairwise to eliminate the quadratic terms, we obtain two linear equations, forming a matrix: A · [ , ] = b Find the least squares solution: [ , ] = (A A) - ¹A b, to obtain the estimated position using triangulation.

[0055] Example continued: Taking the data from the three base stations and the calculated distance above as an example, after elimination and simplification, a system of linear equations is obtained, and the estimated triangulation position is approximately (104.0655, 30.5718).

[0056] (4) Consistency comparison: Calculate the estimated position of triangulation ( , ) and GPS positioning coordinates (x) at the same time gps ,y gps Euclidean distance between them: D dev = √(( - x_gps)² + ( - y_gps)²); Set the error tolerance threshold D threshold =200 meters. If D dev ≤ D threshold If the sampling point is consistent, then it is marked as "consistent"; otherwise, it is marked as "inconsistent".

[0057] Example continued: At this moment, the GPS positioning coordinates are (104.06579, 30.57215), and the estimated position by triangulation is (104.0655, 30.5718). The calculated distance between the two points is approximately 43 meters, which is less than the 200-meter threshold, so they are marked as consistent.

[0058] (5) Sub-score calculation: Calculate the consistency ratio P of all sampling points throughout the entire training period. consistent =N consistent / N, press P consistent The mapping is to sub-score B3. For example, P consistent ≥ 90% gets 1.0 point, 80%-90% gets 0.8 points, 70%-80% gets 0.6 points, and below 70% gets 0.4 points or less.

[0059] Step 4: Score Summary: Combine the above three sub-scores according to a predetermined ratio to form a GPS and base station data consistency score B. B = 0.3 × B1 + 0.4 × B2 + 0.3 × B3.

[0060] B1 is the coverage matching sub-score, B2 is the motion-handover logic detection sub-score, and B3 is the triangulation consistency sub-score. In the weight setting, B2 has the highest weight of 0.4 because its logic detection can best reveal the essence of the simulator's inability to synchronously forge base station handover behavior; B1 and B3 each account for 0.3, providing verification from the two dimensions of static matching and independent location estimation, respectively.

[0061] Step 5: Detection of stationary anomalies For a hypothetical training scenario where the vehicle remains almost stationary throughout the entire training period, the GPS trajectory of a complete training session is captured, and the cumulative displacement of all sampling points is calculated. The cumulative displacement calculation formula is: summing the straight-line distances between all adjacent GPS sampling points segment by segment to obtain the total displacement D. total If D total If the distance is below a preset threshold of 500 meters, and the cumulative mileage increment read from the vehicle's OBD data is also 0, a secondary verification by the base station is initiated: Analysis of all base station information records during this period is performed. If the serving cell CellID remains unchanged, and the absolute value of the difference between the RSSI of each sampling point and the mean RSSI during the period is calculated, if all differences are within a preset decibel range (e.g., ±3dB), it can be determined that the device was indeed placed in a fixed location and did not move with the vehicle, suggesting possible cheating. This result provides strong evidence for the final comprehensive score and anomaly labeling.

[0062] Step 6: Vehicle motion mode fusion verification, combining sensor data A motion authenticity score is generated using accelerometer and gyroscope data, which is then integrated into the vehicle operation data authenticity score C. The amplitude and variation patterns of the three-axis acceleration are analyzed: In real vehicle operation, the Z-axis exhibits continuous slight vibrations, and its acceleration variance should be greater than the typical value under stationary conditions; the X / Y axes show specific waveforms during turning or acceleration / deceleration; and the gyroscope detects a significant change in Z-axis angular velocity during turning. Specifically, if the GPS indicates the vehicle is continuously turning with a directional angle change rate >10° / s, but the mean Z-axis angular velocity data from the gyroscope is <2° / s, then the sensor data is deemed inconsistent with the GPS trajectory, and the corresponding sub-item score in score C is reduced. Score C is derived by combining the OBD vehicle speed / mileage reasonableness and sensor motion consistency, with each factor contributing 0.5.

[0063] Step 7: Calculate the overall credibility score This embodiment uses three core scores: GPS trajectory rationality score A, GPS and base station data consistency score B, and vehicle operation data authenticity score C. Weights are set as follows: w1=0.35, w2=0.45, w3=0.20. The overall reliability score is then calculated as follows: S = 0.35 × A + 0.45 × B + 0.20 × C; The system has a preset credibility threshold of T=0.7. If the calculated S<0.7, the recorded learning hours are considered abnormal.

[0064] Example Explanation: After trainee A completed training, the system scores are as follows: GPS trajectory reasonableness score A = 0.85, the trajectory was basically within the training area, but there were a few jump points; GPS and base station data consistency score B = 0.42, where B1 = 0.85, B2 = 0.3, B3 = 0.7, and the overall B = 0.3 × 0.85 + 0.4 × 0.3 + 0.3 × 0.7 = 0.255 + 0.12 + 0.21 = 0.585, calculated according to the values ​​in the example above; vehicle operation data authenticity score C = 0.90, OBD data is normal. Substituting into the formula: S = 0.35 × 0.85 + 0.45 × 0.585 + 0.20 × 0.90 = 0.2975 + 0.26325 + 0.18 = 0.74075. S=0.74≥0.7, the abnormal threshold has not been triggered yet, but because the B score is low, the system still marks the record as "attention" for manual review.

[0065] Step 8: Exception Handling and Feedback When an anomaly is detected (S<0.7) or other obvious cheating indicators are present, the cloud server automatically generates an anomaly record. This record includes: trainee identity, training period, anomaly type (e.g., "GPS base station movement mismatch," "abnormal stationary location," "excessive triangulation deviation"), overall score S and scores for each sub-item, and key evidence (e.g., captured GPS trajectory map, base station switching anomaly time point markers, comparison chart of triangulation and GPS trajectory, etc.). The monitoring platform 003 receives an alert immediately and displays the record prominently, allowing administrators to retrieve details for manual review and determine whether to accept the training hours.

[0066] Compared to the solution in CN202011031766.8, the technical solution in this embodiment eliminates the need for students to carry mobile terminals, does not rely on Bluetooth connection stability, and completely eliminates data interruptions caused by factors such as student phone battery level, permissions, and compatibility. More importantly, this embodiment directly collects base station information through a mobile communication module, rather than simply obtaining base station positioning coordinates. Instead, it performs logical reconciliation analysis on the motion state reflected by the GPS trajectory and the base station network switching behavior, and introduces multi-base station triangulation to provide independent verification, achieving a leap from "data source selection" to "behavioral pattern cross-verification." Combined with vehicle OBD and motion sensors, a three-dimensional anti-counterfeiting system covering wireless signals, satellite positioning, and vehicle physical status is formed, significantly improving the reliability and practicality of training hour monitoring.

[0067] Second Embodiment This embodiment deploys a lightweight verification algorithm in the main control chip of the vehicle-mounted training device 001. This algorithm performs local real-time GPS geofence detection, vehicle speed and acceleration rationality checks, and simple base station handover frequency monitoring. Specifically, the local algorithm calculates the handover frequency by comparing whether the Cell ID changes between two adjacent base station interactions. When the local algorithm does not detect any suspicious features (such as GPS always being within the geofence, continuous vehicle speed, and base station handover frequency roughly matching GPS speed), the device only uploads brief heartbeat data and key summary information every 60 seconds, significantly reducing bandwidth usage. Once preliminary anomalies are detected, such as GPS speed >15km / h but no base station handover within 10 minutes, or trajectory jumps exceeding 300 meters, the original detailed data for the preceding and following 30 minutes is immediately uploaded to the cloud in batches for deep triangulation verification and comprehensive scoring. This solution effectively saves network bandwidth and cloud computing resources while ensuring monitoring effectiveness.

[0068] Third Embodiment This embodiment, based on any of the above schemes, adds continuous verification of the trainee's identity during the training process. After training begins, the vehicle-mounted training device 001 triggers a facial recognition verification every random time interval, such as 10 to 35 minutes. The screen prompts the trainee to look directly at the camera. If the trainee fails to complete the verification within 60 seconds, or if the verification fails, the training time record for that period is marked as "identity questionable." The cloud-based comprehensive scoring algorithm incorporates the success rate of identity verification into its auxiliary judgment: if multiple verification failures occur, the final credibility score is significantly reduced or the system is directly deemed invalid, thereby preventing cheating by switching trainees mid-training.

[0069] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices or units, or may be electrical, mechanical or other forms of connection.

[0070] The units described as separate components may or may not be physically separate. As will be appreciated by those skilled in the art, the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0071] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for monitoring the authenticity of driver training hours based on dual verification of GPS and base stations, characterized in that, Includes the following steps: Data collection steps: During student training, GPS positioning data, base station information data, and vehicle operation data are continuously collected using onboard training equipment installed in the training vehicle. The base station information data includes at least the cell ID, location area code, signal strength, and a list of neighboring base stations and their signal strengths for the currently connected base station. In this data collection step, the GPS positioning data and base station information data are collected every 10-60 seconds, and the vehicle operation data is collected every 1-10 seconds. The dual verification process involves cross-verifying the GPS positioning data and the base station information data. This cross-verification includes: calculating the matching degree between the GPS positioning location and the theoretical coverage area of ​​the currently connected base station; and determining data anomalies if the GPS positioning data shows vehicle movement while the base station information data indicates no switching, or if the GPS positioning data shows vehicle stationary while the base station information data indicates frequent switching; and performing triangulation calculations using the signal strength of multiple base stations in the adjacent base station list to obtain a location estimate independent of GPS positioning, which is then compared and verified with the GPS positioning location. Location anomaly detection step: Calculate the total displacement obtained from the GPS positioning data within a training period. If the total displacement is lower than a preset displacement threshold and the vehicle operation data indicates no effective mileage, then a secondary confirmation is performed based on the base station information data: If the base station information data indicates that the base station does not switch at all during the training period and the signal strength fluctuation is lower than a preset fluctuation threshold, then it is determined that the vehicle-mounted training equipment is in a fixed position, which is suspected of cheating; In the location anomaly detection step, the preset displacement threshold is 500 meters; The secondary confirmation specifically means: If the base station does not perform any cell handover during the training period and the change in signal strength is always less than a preset decibel value, then it is determined that the equipment is in a fixed position; Multi-dimensional fusion scoring steps: Calculate the comprehensive credibility score S based on a preset multi-dimensional verification index system, wherein the index system includes at least a GPS trajectory rationality score, a GPS and base station data consistency score, and a vehicle operation data authenticity score; and obtain the comprehensive credibility score S by weighted summation of each score. Anomaly handling steps: When the comprehensive credibility score S is lower than the preset credibility threshold, the corresponding learning hour record is determined to be abnormal, the abnormal information and abnormal evidence are recorded, and displayed through the supervision platform for manual review.

2. The method according to claim 1, characterized in that, Before the dual verification step, a data preprocessing step is also included: the collected GPS positioning data, base station information data and vehicle operation data are synchronized and aligned according to the timestamp, and the GPS trajectory data is filtered to remove abnormal positioning points.

3. The method according to claim 1, characterized in that, The dual verification step also includes GPS trajectory rationality verification, which includes: analyzing whether the GPS trajectory is within the preset training site area; calculating the distance and speed between adjacent GPS positioning points to determine whether the speed change is continuous and whether the acceleration is within a reasonable range; and detecting whether the GPS trajectory has abnormal jumps, instantaneous shifts or repetitions.

4. The method according to claim 1, characterized in that, In the multi-dimensional fusion scoring step, the GPS trajectory rationality score is recorded as A, the GPS and base station data consistency score is recorded as B, and the vehicle operation data authenticity score is recorded as C; the formula for calculating the comprehensive credibility score S is: S = w1 × A + w2 × B + w3 × C; Where w1, w2, and w3 are weight coefficients, satisfying w1 + w2 + w3 = 1, and the range of each weight's proportion is: w1∈ [0.3, 0.4], w2∈ [0.4, 0.5], w3∈ [0.2, 0.3].

5. The method according to claim 1, characterized in that, The method further includes: deploying a lightweight verification algorithm on the vehicle-mounted training device to perform preliminary anomaly screening on the collected data, and uploading relevant detailed data to the cloud verification server for in-depth analysis only when the preliminary anomaly screening detects suspicious situations.

6. The method according to claim 1, characterized in that, The vehicle operation data also includes vehicle motion pattern data collected by accelerometers and gyroscopes integrated in the vehicle training equipment; in the multi-dimensional fusion scoring step, a score is also given based on the consistency between the vehicle motion pattern data and the GPS positioning data, in order to participate in the calculation of the comprehensive credibility score.

7. The method according to claim 1, characterized in that, The method also includes a continuous biometric authentication step: at the start of training, the trainee is initially authenticated, and during the training process, multiple identity verifications are randomly initiated through the biometric recognition module integrated into the vehicle-mounted training equipment. The verification results are used to assist in judging the authenticity of the training hour records.

8. A driver training hour authenticity monitoring system based on dual verification of GPS and base stations, characterized in that, include: The vehicle-mounted training equipment, installed on the training vehicle, integrates a GPS positioning module and a mobile communication module. It is used to collect GPS positioning data, base station information data, and vehicle operation data during the training period and upload them through the mobile communication network. The base station information data includes at least the cell ID, location area code, signal strength, and a list of adjacent base stations and their signal strengths of the currently connected base station. A cloud-based verification server, connected to the network of the vehicle-mounted training equipment, is used to receive the data and execute the method as described in any one of claims 1 to 7; The monitoring platform is connected to the cloud verification server via a network and is used to receive and display study hour records, abnormal alarms, and data statistics.

9. The system according to claim 8, characterized in that, The vehicle-mounted training equipment also integrates an accelerometer and a gyroscope for collecting vehicle motion pattern data; when executing the method, the cloud verification server also uses the vehicle motion pattern data for fusion verification.

10. The system according to claim 8, characterized in that, The vehicle-mounted training equipment also integrates a biometric recognition module for randomly verifying the identity of trainees multiple times during the training process; when executing the method, the cloud verification server also combines the identity verification results to determine the validity of the training hours.

Images