Commercial vehicle abnormal refueling infrared linkage identification and tracking system and method

Abstract

The application provides a commercial vehicle abnormal refueling infrared linkage identification and tracing system and method. The system comprises a liquid level change sensing module, a geographic fence intelligent filtering module, an abnormality determination module, an infrared video linkage evidence extraction module and an evidence chain generation module. The method comprises the following sub-steps: S1, identifying a refueling event and recording a timestamp; S2, acquiring the latitude and longitude of the vehicle; S3, performing three-level compliance filtering to determine whether it is abnormal; S4, extracting and solidifying video evidence; and S5, generating an electronic evidence package containing multi-source data. The method collects fuel liquid level data through a vehicle-mounted sensor, uploads and analyzes the data in real time, eliminates the fraud space caused by manual operation from the source, eliminates the need for drivers to manually submit refueling receipts, reduces the human intervention link, sets multiple determination standards and determines real parking events and abnormal refueling events in real time, realizes the real-time response, avoids the subjective randomness of manual review, and improves the determination accuracy.

Description

Technical Field

[0001] This invention relates to the field of intelligent transportation and vehicle management technology, and in particular to an infrared linkage identification and traceability system and method for abnormal refueling of commercial vehicles. Background Technology

[0002] In commercial vehicle operation and management, fuel cost is one of the core operating expenses, usually accounting for more than 30% of the total operating cost, which directly affects the company's profitability. At the same time, fuel quality is closely related to vehicle operation safety and service life. Therefore, companies generally require vehicles to refuel only at designated brand gas stations (such as the "three barrels of oil": Sinopec, PetroChina, and CNOOC) or authorized stations in order to achieve the dual goals of cost control and fuel quality assurance. With the intelligent upgrade of commercial vehicles, commercial vehicles meeting the China V emission standard and above are generally equipped with fuel tank level sensors and on-board intelligent terminals (T-BOX), which have the functions of fuel level change monitoring and location reporting. However, relying solely on the simple judgment logic of "fuel level rise + GPS coordinates" to determine whether a violation has occurred has the following drawbacks: The manual supervision model is inefficient and easy to falsify: Traditional fuel control relies on drivers to manually fill in refueling information and submit refueling receipts for review. It is entirely dependent on manual operation. Drivers can evade supervision by forging receipts, falsely reporting refueling amounts or locations, etc. In addition, manual review cannot achieve real-time intervention. By the time problems are discovered, economic losses have already occurred, and the timeliness and effectiveness of supervision are seriously insufficient. Hidden illegal refueling behavior is difficult to identify: Some drivers, in order to seek personal gain, take advantage of existing technical loopholes to choose to refuel at non-compliant sites such as nameless small stations or private mobile tanker trucks at night or on remote roads. They then sell the fuel quota allocated by the company at a low price, while creating the illusion of compliant refueling. Based on the existing judgment logic alone, it is not possible to accurately identify this illegal behavior. Broken chain of evidence makes accountability difficult: Even if anomalies are discovered, the lack of existing technology for collecting on-site evidence that can corroborate violations makes it difficult to hold anyone accountable. Summary of the Invention

[0003] The purpose of this invention is to address the shortcomings of existing technologies by proposing an infrared linkage identification and traceability system and method for abnormal refueling of commercial vehicles.

[0004] To achieve the above objectives, the present invention adopts the following technical solution: The commercial vehicle abnormal refueling infrared linkage identification and traceability system includes a liquid level change sensing module, a geofence intelligent filtering module, an anomaly judgment module, an infrared video linkage evidence collection module, and an evidence chain generation module. The liquid level change sensing module is used to collect fuel level data and vehicle data, determine real parking events, and trigger the generation of refueling events. The geofence intelligent filtering module is used to obtain the latitude and longitude of the vehicle in the refueling event and perform validity verification. After obtaining the structured address through reverse geocoding, it performs three-level compliance filtering to screen out suspected abnormal refueling events and transmit the relevant full data. The anomaly determination module is used to make a final determination of abnormal refueling events; The infrared video linkage evidence collection module is used to extract video clips from preset time windows before and after abnormal refueling events, add digital watermarks, and solidify them as video evidence. The evidence chain generation module is used to generate electronic evidence packages.

[0005] This invention also proposes a method for infrared linkage identification and traceability of abnormal refueling in commercial vehicles, including the following sub-steps: S1: Identify refueling events and record timestamps; Includes the following sub-steps: S11: Collect fuel level data and upload it to the on-board intelligent terminal of the commercial vehicle; The fuel level change sensing module acquires real-time fuel level data in the fuel tank through a built-in fuel level sensor in the commercial vehicle's fuel tank. The fuel level data includes the percentage of fuel level in the tank and a timestamp. The fuel level data is then uploaded to the commercial vehicle's onboard intelligent terminal in real time via an onboard data transmission protocol. S12: Commercial vehicle in-vehicle intelligent terminal synchronously collects in-vehicle data of commercial vehicles; The on-board intelligent terminal of the commercial vehicle receives fuel level data and simultaneously collects on-board data of the commercial vehicle, including vehicle speed, odometer mileage, license plate number and corresponding timestamp; the fuel level data and on-board data are saved to the local database in chronological order of timestamps; S13: Determine whether it is a real parking event based on vehicle data; The liquid level change sensing module is pre-set with real parking event standards. It compares the vehicle speed, instrument mileage and other vehicle data collected by the commercial vehicle's on-board intelligent terminal with the real parking event standards. If the real parking event standards are met, it is determined to be a real parking event. The start and end timestamps of the real parking event are recorded and the process proceeds to step S14. Otherwise, it is determined to be a non-real parking event. S14: Perform stability verification on fuel level data; The liquid level change sensing module is set to a duration t, where t ≥ 3 minutes; Based on the start and end timestamps of the actual parking events, two sets of fuel level data are extracted from the local database, namely, data segment A1 and data segment A2. Data segment A1 consists of all fuel level data during the stable driving state of the vehicle within a time period t before the start timestamp of the actual parking event; data segment A2 consists of all fuel level data during the stable driving state of the vehicle within a time period t after the end timestamp of the actual parking event. The fuel level data is then subjected to a stability check. If the stability check passes, proceed to step S15; otherwise, the stability check of the fuel level data is deemed to have failed, and the process terminates. S15: Determine the refueling incident; The fuel level change sensing module extracts the fuel level percentage of the fuel tank that is closest to the start timestamp of the actual parking event from the A1 segment data, and denots it as L1; it extracts the fuel level percentage of the fuel tank that is closest to the end timestamp of the actual parking event from the A2 segment data, and denots it as L2. When |L1-L2|>10%, it is determined to be a refueling event and a unique event identifier is generated. The liquid level change sensing module sends a refueling event trigger signal and a unique event identifier to the commercial vehicle's on-board intelligent terminal; otherwise, it is determined to be a normal fluctuation in liquid level and is not a refueling event.

[0006] S2: Obtain vehicle latitude and longitude; The onboard intelligent terminal of the commercial vehicle receives the refueling event trigger signal and the unique event identifier, and binds the unique event identifier, timestamp and the corresponding license plate number; through the built-in GPS and Beidou dual-mode positioning chip, it obtains the vehicle's latitude and longitude coordinates at the timestamp in real time based on the license plate number.

[0007] S3: Perform three-level compliance filtering to determine if there are any anomalies; S31: Verify the validity of the vehicle's latitude and longitude coordinates; The geofence intelligent filtering module verifies the validity of the acquired vehicle latitude and longitude coordinates, checking whether the vehicle latitude and longitude are within the preset range. If not, the coordinates are determined to be invalid and an error is indicated; if they are, the process proceeds to step S32. S32: Perform reverse geocoding on the vehicle's latitude and longitude coordinates to obtain the coding result; The geofencing intelligent filtering module calls upon a cloud-based reverse geocoding engine, which is associated with an electronic map vector database. This database contains actual geographic points, each with corresponding structured address information. The structured address information includes administrative level information, road and geographic landmark information, location type, and gas station brand information, among other things. The vehicle's latitude and longitude coordinates and the corresponding unique event identifier are transmitted as input parameters to the reverse geocoding engine. The reverse geocoding engine uses a spatial distance matching algorithm in the electronic map vector database to filter out the geospatial point that is closest to the vehicle's latitude and longitude coordinates and retrieves the structured address information corresponding to the geospatial point. The structured address information, vehicle latitude and longitude coordinates, and unique event identifier are bound together as the encoding result; and the encoding result is sent back to the geofence intelligent filtering module. S33: Perform three-level compliance filtering on the structured address information in the encoding results; Includes the following sub-steps: S331: Perform brand whitelist filtering; The geofencing intelligent filtering module compares the structured address information in the encoding results with a pre-set list of compliant gas station brands and a whitelist of registered addresses. The list of compliant gas station brands contains compliant gas station brands, and the whitelist of registered addresses contains compliant gas station addresses. If the structured address information contains any compliant gas station brand from the list of compliant gas station brands or any compliant gas station address from the whitelist of registered addresses, then If the condition is not met, it is determined that the refueling is normal; if not, it is determined that the brand whitelist filter has not been passed, and proceed to step S332. S332: Performs high-speed scene filtering; The geofence intelligent filtering module performs keyword retrieval on the structured address information. If the structured address information contains keywords such as highway or highway service area, it is determined to be normal refueling; if not, it is determined to have failed the highway scene filtering and proceeds to step S333. S333: Conduct backup filtration at compliant gas stations; Pre-set the search radius; The geofencing intelligent filtering module uses the vehicle's latitude and longitude coordinates as the center. Within the search radius, it searches the official gas station POI database for gas station sites. If a compliant gas station brand is found, it is determined to be a normal refueling; otherwise, it is determined to be a suspected abnormal event that has not passed the compliant gas station fallback filter. The full data of the suspected abnormal event is then transmitted to the anomaly judgment module. The full data of the suspected abnormal event includes the vehicle's latitude and longitude coordinates, a unique event identifier, the coding result, and the relevant judgment results of the three-level compliance filter. S34: Final determination of compliance filtering at level three; The anomaly detection module performs a final judgment on all data of suspected abnormal events. If the event fails to pass the brand whitelist filter, the highway scenario filter, and the compliant gas station fallback filter, it is ultimately determined to be an abnormal refueling event, and a unique event ID and corresponding timestamp are generated. Bind all data of abnormal refueling events with the event's unique ID and corresponding timestamp; The complete data for the abnormal refueling event includes the vehicle's latitude and longitude coordinates, the event's unique identifier, the coding result, and the relevant judgment results of the three-level compliance filtering.

[0008] S4: Extract and preserve video evidence; After the cloud determines that an abnormal refueling event has occurred, it sends a video saving instruction to the vehicle gateway. The video saving instruction includes the unique event ID of the abnormal event, the timestamp, and the preset video event window. The vehicle gateway receives the instruction and retrieves the video clips within a preset time window before and after the timestamp corresponding to the abnormal event through the infrared video linkage evidence collection module. The video is collected in real time by the vehicle infrared camera device that is always in a state of continuous recording with constant power. The infrared video linkage evidence collection module adds digital watermarks to the extracted video clips. The digital watermarks include vehicle latitude and longitude coordinates, license plate number, timestamp, etc., and are solidified as video evidence. S5: Generate an electronic evidence package containing multi-source data; The evidence chain generation module associates and encapsulates all data from the abnormal refueling event with video evidence, generates an electronic evidence package, and transmits it to the cloud database.

[0009] Compared with the prior art, the beneficial effects of the present invention are as follows: This method collects fuel level data and vehicle data through onboard sensors and commercial vehicle intelligent terminals, uploads and analyzes the data in real time, eliminating the possibility of fraud caused by manual operation at the source. It eliminates the need for drivers to manually submit refueling receipts, reducing human intervention. It sets multiple judgment criteria and judges real parking events and abnormal refueling events in real time, achieving real-time response, avoiding the subjectivity and arbitrariness of manual review, and improving the accuracy of judgment.

[0010] By combining reverse geocoding with three-level compliance filtering, integrating multi-dimensional matching of brand, scenario, and geographical range, and combining spatial distance matching algorithm with official POI database for surrounding verification, it breaks through the limitations of traditional single GPS coordinate judgment, realizes refined and full-dimensional verification of refueling location, and solves the problem of identifying non-compliant refueling in the vicinity where the coordinates do not directly fall on the gas station. By using both data and video evidence, the evidence is solidified and tamper-proofed, and multi-source data is packaged and uploaded to the cloud, forming a complete and traceable chain of evidence, providing effective evidence for accountability. Attached Figure Description

[0011] Figure 1 This is a flowchart illustrating the steps of the infrared linkage identification and traceability method for abnormal refueling of commercial vehicles according to the present invention. Detailed Implementation

[0012] To provide a further understanding of the purpose, structure, features, and functions of the present invention, detailed descriptions are provided below with reference to specific embodiments.

[0013] The commercial vehicle abnormal refueling infrared linkage identification and traceability system includes a liquid level change sensing module, a geofence intelligent filtering module, an anomaly judgment module, an infrared video linkage evidence collection module, and an evidence chain generation module. The liquid level change sensing module is used to collect fuel level data and vehicle data, determine real parking events, and trigger the generation of refueling events. The geofence intelligent filtering module is used to obtain the latitude and longitude of the vehicle in the refueling event and perform validity verification. After obtaining the structured address through reverse geocoding, it performs three-level compliance filtering to screen out suspected abnormal refueling events and transmit the relevant full data. The anomaly determination module is used to make a final determination of abnormal refueling events; The infrared video linkage evidence collection module is used to extract video clips from preset time windows before and after abnormal refueling events, add digital watermarks, and solidify them as video evidence. The evidence chain generation module is used to generate electronic evidence packages.

[0014] like Figure 1 As shown, this invention also proposes a method for infrared linkage identification and traceability of abnormal refueling in commercial vehicles, including the following sub-steps: S1: Identify refueling events and record timestamps; Includes the following sub-steps: S11: Collect fuel level data and upload it to the on-board intelligent terminal of the commercial vehicle; The fuel level change sensing module acquires fuel level data in real time through a fuel level sensor built into the commercial vehicle's fuel tank. The fuel level data includes the percentage of fuel level in the tank and a timestamp. The fuel level data is then uploaded to the commercial vehicle's onboard intelligent terminal (T-BOX) in real time via an onboard data transmission protocol (such as the O200 protocol). Specifically, a mapping relationship between the fuel tank volume and the fuel level is pre-set. The fuel tank volume is known. The fuel level sensor collects the fuel level in the fuel tank in real time and converts the fuel level into the actual fuel volume according to the known mapping relationship between the fuel tank volume and the fuel level. The proportion of the fuel volume to the fuel tank volume is the percentage of the fuel level in the fuel tank. S12: Commercial vehicle in-vehicle intelligent terminal synchronously collects in-vehicle data of commercial vehicles; The on-board intelligent terminal of the commercial vehicle receives fuel level data and simultaneously collects on-board data of the commercial vehicle, including vehicle speed, odometer mileage, license plate number and corresponding timestamp; the fuel level data and on-board data are saved to the local database in chronological order of timestamps; S13: Determine whether it is a real parking event based on vehicle data; The liquid level change sensing module is pre-set with real parking event standards. It compares the vehicle speed, instrument mileage and other vehicle data collected by the commercial vehicle's on-board intelligent terminal with the real parking event standards. If the real parking event standards are met, it is determined to be a real parking event. The start and end timestamps of the real parking event are recorded and the process proceeds to step S14. Otherwise, it is determined to be a non-real parking event. For example, the standard for a real parking event is: a real parking event is defined as a situation where the vehicle speed is 0 for ≥3 minutes and the change in odometer mileage is ≤2km during that period. If a vehicle's speed is 0 for less than 3 minutes due to reasons such as waiting at a red light, temporarily pulling over, or yielding to pedestrians, it is considered a short stop and not a true parking event. If the vehicle speed is 0 but the odometer reading changes by more than 2km during that period, it is determined to be a non-real parking event. S14: Perform stability verification on fuel level data; The liquid level change sensing module is set to a duration t, where t ≥ 3 minutes; Based on the start and end timestamps of real parking events, two sets of fuel level data are extracted from the local database, namely, data segment A1 and data segment A2. The data segment A1 consists of all fuel level data under stable driving conditions within a time period t before the start timestamp of the real parking event; the data segment A2 consists of all fuel level data under stable driving conditions within a time period t after the end timestamp of the real parking event; and the stability of the fuel level data is verified. The specific stability verification method is as follows: The arithmetic mean algorithm is used to obtain the mean percentage of fuel level in the tank for data in segments A1 and A2, respectively, which are denoted as the mean of fuel level in segment A1 and the mean of fuel level in segment A2. Extract the fuel level percentage of the last fuel tank in segment A1 and the fuel level percentage of the last fuel tank in segment A2; The percentage deviations for segment A1 and segment A2 are obtained using the following formulas: A1 segment deviation percentage = (|Average A1 segment level - Percentage of fuel level in the last tank of A1 segment| ÷ Average A1 segment level) × 100%; A2 segment deviation percentage = (|Average A2 segment level - Percentage of fuel level in the last tank of A2 segment| ÷ Average A2 segment level) × 100%; If the deviation percentage of segment A1 is ≤5% and the deviation percentage of segment A2 is ≤5%, the stability verification of the fuel level data is deemed to have passed, and the process proceeds to step S15; otherwise, the stability verification of the fuel level data is deemed to have failed. S15: Determine the refueling incident; The fuel level change sensing module extracts the fuel level percentage of the fuel tank that is closest to the start timestamp of the actual parking event from the A1 segment data, and denots it as L1; it extracts the fuel level percentage of the fuel tank that is closest to the end timestamp of the actual parking event from the A2 segment data, and denots it as L2. When |L1-L2|>10%, it is determined to be a refueling event and a unique event identifier is generated. The liquid level change sensing module sends a refueling event trigger signal and a unique event identifier to the commercial vehicle's on-board intelligent terminal; otherwise, it is determined to be a normal fluctuation in liquid level and is not a refueling event.

[0015] By employing a dual pre-verification mechanism that combines real-world parking event assessment with fuel level stability verification, interference factors such as short-term parking and fuel level data fluctuations are eliminated before refueling events are determined. This solves the misjudgment problem caused by relying solely on fuel level rises in traditional technologies. Furthermore, the stability of the data is verified by the percentage deviation (≤5%) between the mean and end values ​​of the A1 / A2 segment data, ensuring that the fuel level data used to determine refueling events is authentic and valid, thus improving data reliability.

[0016] S2: Obtain vehicle latitude and longitude; The onboard intelligent terminal of the commercial vehicle receives the refueling event trigger signal and the unique event identifier, and binds the unique event identifier, timestamp and the corresponding license plate number; through the built-in GPS and Beidou dual-mode positioning chip, it obtains the vehicle's latitude and longitude coordinates at the timestamp in real time based on the license plate number.

[0017] S3: Perform three-level compliance filtering to determine if there are any anomalies; S31: Verify the validity of the vehicle's latitude and longitude coordinates; The geofence intelligent filtering module verifies the validity of the acquired vehicle latitude and longitude coordinates, checking whether the vehicle latitude and longitude are within the preset range. If not, the coordinates are determined to be invalid and an error is indicated; if they are, the process proceeds to step S32. S32: Perform reverse geocoding on the vehicle's latitude and longitude coordinates to obtain the coding result; The geofencing intelligent filtering module calls upon a cloud-based reverse geocoding engine, which is associated with an electronic map vector database. This database contains actual geographic points, each with corresponding structured address information. The structured address information includes administrative level information, road and geographic landmark information, location type, and gas station brand information, among other things. Specifically, the administrative level information refers to provinces / autonomous regions / municipalities, cities / prefectures, districts / counties, townships / towns / streets, etc.; the road and geographical identification information refers to the national highway / provincial highway / county road number and name (such as G308, X212), surrounding villages / industrial parks / landmark buildings, etc.; the location type and brand information refers to the location category of the entity corresponding to the address (such as gas station, highway service area, residential area, etc.), and if it is a gas station / service area, it includes its specific brand name (such as China National Petroleum Corporation, Sinopec, etc.). The vehicle's latitude and longitude coordinates and the corresponding unique event identifier are transmitted as input parameters to the reverse geocoding engine. The reverse geocoding engine uses a spatial distance matching algorithm in the electronic map vector database to filter out the geospatial point that is closest to the vehicle's latitude and longitude coordinates and retrieves the structured address information corresponding to the geospatial point. The structured address information, vehicle latitude and longitude coordinates, and unique event identifier are bound together as the encoding result; and the encoding result is sent back to the geofence intelligent filtering module. S33: Perform three-level compliance filtering on the structured address information in the encoding results; Includes the following sub-steps: S331: Perform brand whitelist filtering; The geofencing intelligent filtering module compares the structured address information in the encoding results with a pre-set list of compliant gas station brands and a whitelist of registered addresses. The list of compliant gas station brands contains compliant gas station brands, and the whitelist of registered addresses contains compliant gas station addresses. If the structured address information contains any compliant gas station brand from the list of compliant gas station brands or any compliant gas station address from the whitelist of registered addresses, then If the condition is not met, it is determined that the refueling is normal; if not, it is determined that the brand whitelist filter has not been passed, and proceed to step S332. S332: Performs high-speed scene filtering; The geofence intelligent filtering module performs keyword retrieval on the structured address information. If the structured address information contains keywords such as highway or highway service area, it is determined to be normal refueling; if not, it is determined to have failed the highway scene filtering and proceeds to step S333. S333: Conduct backup filtration at compliant gas stations; Pre-set the search radius; The geofencing intelligent filtering module uses the vehicle's latitude and longitude coordinates as the center. Within the search radius, it searches the official gas station POI database for gas station sites. If a compliant gas station brand is found, it is determined to be a normal refueling; otherwise, it is determined to be a suspected abnormal event that has not passed the compliant gas station fallback filter. The full data of the suspected abnormal event is then transmitted to the anomaly judgment module. The full data of the suspected abnormal event includes the vehicle's latitude and longitude coordinates, a unique event identifier, the coding result, and the relevant judgment results of the three-level compliance filter. S34: Final determination of compliance filtering at level three; The anomaly detection module performs a final judgment on all data of suspected abnormal events. If the event fails to pass the brand whitelist filter, the highway scenario filter, and the compliant gas station fallback filter, it is ultimately determined to be an abnormal refueling event, and a unique event ID and corresponding timestamp are generated. Bind all data of abnormal refueling events with the event's unique ID and corresponding timestamp; The complete data for the abnormal refueling event includes the vehicle's latitude and longitude coordinates, the event's unique identifier, the coding result, and the relevant judgment results of the three-level compliance filtering.

[0018] Through a three-level filtering system, it considers both the compliant brands / addresses designated by enterprises and special compliant scenarios such as highway service areas. It also uses a POI database to conduct a fallback search of nearby compliant gas stations, thoroughly investigating non-compliant gas station scenarios and preventing violations from exploiting loopholes. Combined with GPS + Beidou dual-mode positioning to obtain vehicle latitude and longitude, the system achieves higher positioning accuracy and stronger anti-interference capabilities. At the same time, it verifies the validity of latitude and longitude coordinates (latitude and longitude within a preset geographical range) to prevent misjudgments caused by falsified or invalid positioning data, ensuring the accuracy of location determination.

[0019] S4: Extract and preserve video evidence; After the cloud determines that an abnormal refueling event has occurred, it sends a video saving instruction to the vehicle gateway. The video saving instruction includes the unique event ID of the abnormal event, the timestamp, and a preset video event window (60 seconds before and after the timestamp corresponding to the abnormal refueling event). The vehicle gateway receives the instruction and retrieves the video clips within a preset time window before and after the timestamp corresponding to the abnormal event through the infrared video linkage evidence collection module. The video is collected in real time by the vehicle infrared camera device that is always in a state of continuous recording with constant power. The infrared video linkage evidence collection module adds digital watermarks to the extracted video clips. The digital watermarks include vehicle latitude and longitude coordinates, license plate number, timestamp, etc., and are solidified as video evidence. S5: Generate an electronic evidence package containing multi-source data; The evidence chain generation module associates and encapsulates all data from the abnormal refueling event with video evidence, generates an electronic evidence package, and transmits it to the cloud database.

[0020] By adding digital watermarks containing latitude and longitude, license plate number, and timestamp to video evidence and storing them securely, the authenticity and legality of the evidence are ensured, preventing tampering and forgery, and can be directly used as valid evidence for accountability. The evidence chain generation module associates all digital data with the secured video evidence, encapsulates it into an electronic evidence package, and uploads it to a cloud database for permanent storage. This ensures the evidence is traceable and searchable, solving the problems of fragmented, easily lost, and untraceable evidence in traditional methods.

[0021] The present invention has been described in the above-described embodiments; however, these embodiments are merely examples for implementing the present invention. It must be noted that the disclosed embodiments do not limit the scope of the present invention. Conversely, any modifications and refinements made without departing from the spirit and scope of the present invention are within the scope of patent protection of the present invention.

Claims

1. A method for infrared linkage identification and traceability of abnormal refueling in commercial vehicles, characterized in that: Includes the following steps: S1: Identify refueling events and record timestamps; S11: Collect fuel level data and upload it to the on-board intelligent terminal of the commercial vehicle; S12: Commercial vehicle in-vehicle intelligent terminal synchronously collects in-vehicle data of commercial vehicles; S13: Determine whether it is a real parking event based on vehicle data; S14: Perform stability verification on fuel level data; S15: Determine the refueling incident; S2: Obtain vehicle latitude and longitude; S3: Perform three-level compliance filtering to determine if there are any anomalies; S31: Verify the validity of the vehicle's latitude and longitude coordinates; S32: Perform reverse geocoding on the vehicle's latitude and longitude coordinates to obtain the coding result; S33: Perform three-level compliance filtering on the structured address information in the encoding results; S34: Final determination of compliance filtering at level three; S4: Extract and preserve video evidence; Generate full data on abnormal refueling events and solidify video evidence; S5: Generate an electronic evidence package containing multi-source data; The evidence chain generation module associates and encapsulates all data from the abnormal refueling event with video evidence, generates an electronic evidence package, and transmits it to the cloud database.

2. The infrared linkage identification and traceability method for abnormal refueling of commercial vehicles as described in claim 1, characterized in that: The specific details of step S1 are as follows: S11: Collect fuel level data and upload it to the on-board intelligent terminal of the commercial vehicle; The fuel level change sensing module acquires real-time fuel level data in the fuel tank through a built-in fuel level sensor in the commercial vehicle's fuel tank. The fuel level data includes the percentage of fuel level in the tank and a timestamp. The fuel level data is then uploaded to the commercial vehicle's onboard intelligent terminal in real time via an onboard data transmission protocol. S12: Commercial vehicle in-vehicle intelligent terminal synchronously collects in-vehicle data of commercial vehicles; The on-board intelligent terminal of the commercial vehicle receives fuel level data and simultaneously collects on-board data of the commercial vehicle, including vehicle speed, odometer reading, license plate number and corresponding timestamp; Fuel level data and vehicle data are saved to the local database in chronological order according to timestamps. S13: Determine whether it is a real parking event based on vehicle data; The liquid level change sensing module is pre-set with real parking event standards. It compares the vehicle data collected by the commercial vehicle's on-board intelligent terminal with the real parking event standards. If the real parking event standards are met, it is determined to be a real parking event. The start and end timestamps of the real parking event are recorded and the process proceeds to step S14. Conversely, it is determined to be a non-genuine parking event; S14: Perform stability verification on fuel level data; The fuel level change sensing module performs stability verification on the fuel level data. If the stability verification passes, proceed to step S15; otherwise, it determines that the stability verification of the fuel level data fails and terminates the process. S15: Determine the refueling incident; The fuel level change sensing module extracts the fuel level percentage of the fuel tank that is closest to the start timestamp of the actual parking event from the A1 segment data, and denots it as L1; it extracts the fuel level percentage of the fuel tank that is closest to the end timestamp of the actual parking event from the A2 segment data, and denots it as L2. When |L1-L2|>10%, it is determined to be a refueling event and a unique event identifier is generated. The liquid level change sensing module sends a refueling event trigger signal and a unique event identifier to the commercial vehicle's on-board intelligent terminal; otherwise, it is determined to be a normal fluctuation in liquid level and is not a refueling event.

3. The infrared linkage identification and traceability method for abnormal refueling of commercial vehicles as described in claim 1, characterized in that: In step S2, the on-board intelligent terminal of the commercial vehicle receives the refueling event trigger signal and the unique event identifier, and binds the unique event identifier, timestamp and corresponding license plate number; Using a built-in GPS and Beidou dual-mode positioning chip, the vehicle's latitude and longitude coordinates at the timestamp are obtained in real time based on the license plate number.

4. The infrared linkage identification and traceability method for abnormal refueling of commercial vehicles as described in claim 1, characterized in that: The specific details of step S3 are as follows: S31: Verify the validity of the vehicle's latitude and longitude coordinates; The geofence intelligent filtering module verifies the validity of the acquired vehicle latitude and longitude coordinates, checking whether the vehicle latitude and longitude are within a preset range. If not, the coordinates are determined to be invalid and an error is indicated. If present, proceed to step S32; S32: Perform reverse geocoding on the vehicle's latitude and longitude coordinates to obtain the coding result; The geofencing intelligent filtering module calls upon a cloud-based reverse geocoding engine, which is associated with an electronic map vector database. This database contains actual geographic points, each with corresponding structured address information. The structured address information includes administrative level information, road and geographic landmark information, location type, and gas station brand information. The vehicle's latitude and longitude coordinates and the corresponding unique event identifier are transmitted as input parameters to the reverse geocoding engine. The reverse geocoding engine uses a spatial distance matching algorithm in the electronic map vector database to filter out the geospatial point that is closest to the vehicle's latitude and longitude coordinates and retrieves the structured address information corresponding to the geospatial point. The structured address information, vehicle latitude and longitude coordinates, and unique event identifier are bound together as the encoding result; The encoding results are then sent back to the geofence intelligent filtering module. S33: Perform three-level compliance filtering on the structured address information in the encoding results; S331: Perform brand whitelist filtering; S332: Performs high-speed scene filtering; S333: Conduct backup filtration at compliant gas stations; S34: Final determination of compliance filtering at level three; The anomaly detection module performs a final judgment on all data of suspected abnormal events. If the event fails to pass the three-level compliance filter, it is ultimately determined to be an abnormal refueling event, and a unique event ID and corresponding timestamp are generated. Bind all data of abnormal refueling events with the event's unique ID and corresponding timestamp; The complete data for the abnormal refueling event includes the vehicle's latitude and longitude coordinates, the event's unique identifier, the coding result, and the relevant judgment results of the three-level compliance filtering.

5. The infrared linkage identification and traceability method for abnormal refueling of commercial vehicles as described in claim 1, characterized in that: In step S4, after the cloud determines that it is an abnormal refueling event, it sends a video saving instruction to the vehicle gateway. The video saving instruction includes the unique event ID of the abnormal event, the timestamp, and the preset video event window. The vehicle gateway receives the instruction and retrieves the video clips within a preset time window before and after the timestamp corresponding to the abnormal event through the infrared video linkage evidence collection module. The video is collected in real time by the vehicle infrared camera device that is always in a state of continuous recording with constant power. The infrared video-linked evidence collection module adds digital watermarks to the extracted video clips. The digital watermarks include the vehicle's latitude and longitude coordinates, license plate number, and timestamp, thus solidifying them as video evidence.

6. The infrared linkage identification and traceability method for abnormal refueling of commercial vehicles as described in claim 4, characterized in that: The specific details of step S33 are as follows: S331: Perform brand whitelist filtering; The geofencing intelligent filtering module compares the structured address information in the encoding results with a pre-set list of compliant gas station brands and a whitelist of registered addresses. The list of compliant gas station brands contains compliant gas station brands, and the whitelist of registered addresses contains compliant gas station addresses. If the structured address information contains any compliant gas station brand from the list of compliant gas station brands or any compliant gas station address from the whitelist of registered addresses, then If the condition is not met, it is determined that the refueling is normal; if not, it is determined that the brand whitelist filter has not been passed, and proceed to step S332. S332: Performs high-speed scene filtering; The geofence intelligent filtering module performs keyword retrieval on the structured address information. If the structured address information contains keywords such as "highway" or "highway service area", it is determined to be normal refueling; if not, it is determined to have failed the highway scene filtering and proceeds to step S333. S333: Conduct backup filtration at compliant gas stations; Pre-set the search radius; The geofencing intelligent filtering module uses the vehicle's latitude and longitude coordinates as the center. Within the search radius, it searches the official gas station POI database for gas station sites. If a compliant gas station brand is found, it is determined to be a normal refueling; otherwise, it is determined to be a suspected abnormal event that has not passed the compliant gas station fallback filter. The full data of the suspected abnormal event is then transmitted to the anomaly determination module. The full data of the suspected abnormal event includes the vehicle's latitude and longitude coordinates, a unique event identifier, the coding result, and the relevant determination results of the three-level compliance filter.

7. A commercial vehicle abnormal refueling infrared linkage identification and traceability system for implementing the method of any one of claims 1-6, characterized in that: It includes a liquid level change sensing module, a geofence intelligent filtering module, an anomaly detection module, an infrared video linkage evidence collection module, and an evidence chain generation module; The liquid level change sensing module is used to collect fuel level data and vehicle data, determine real parking events, and trigger the generation of refueling events. The geofence intelligent filtering module is used to obtain the latitude and longitude of the vehicle in the refueling event and perform validity verification. After obtaining the structured address through reverse geocoding, it performs three-level compliance filtering to screen out suspected abnormal refueling events and transmit the relevant full data. The anomaly determination module is used to make a final determination of abnormal refueling events; The infrared video linkage evidence collection module is used to extract video clips from preset time windows before and after abnormal refueling events, add digital watermarks, and solidify them as video evidence. The evidence chain generation module is used to generate electronic evidence packages.

Images