Method for automatically identifying abnormal refueling event based on refueling scene

Abstract

The invention provides a working method of an abnormal event identification method based on a refueling scene. The working method comprises the following steps: S1, collecting data through a sensor network and a communication module; s2, uploading the collected data to a vehicle-mounted terminal device through a communication module; s3, the vehicle-mounted terminal equipment preprocesses the received data; s4, the parking state of the vehicle is judged, and intervals are divided; s5, detecting the liquid level change of the oil quantity in the oil tank of the vehicle; s6, verifying and outputting three events of oil stealing, oiling and normality; through deep combination of hardware and an algorithm, real-time monitoring of the whole refueling process is realized based on the Internet of Things, a sensor technology and edge calculation, and real-time information of the whole refueling process is sent to a user side platform in time.

Description

Technical Field

[0001] The present invention relates to the field of the Internet of Things, and particularly to a method for automatically identifying abnormal refueling events based on a refueling scenario. Background Art

[0002] With the rapid development of the logistics industry, fleet management has become an important means to optimize operating costs. Among them, refueling management, as an important part of the logistics fleet, strengthening refueling management is an important means to optimize operating costs.

[0003] The logistics industry generally faces the following refueling management problems: 1. High cost of manual monitoring: Traditional refueling management relies on manual records and supervision, with problems of low efficiency and easy errors.

[0004] 2. Severe data island phenomenon: Vehicle driving data, gas station transaction data, and fuel consumption data are difficult to be unified and integrated for analysis.

[0005] 3. Lag in identifying abnormal events: Abnormal events including unauthorized refueling, false reporting of refueling volume, and fuel quality problems are difficult to be discovered in a timely manner.

[0006] 4. Lack of real-time and automated means: Existing solutions are often limited to post-event statistical analysis and cannot achieve real-time monitoring and processing. Summary of the Invention

[0007] The purpose of the present invention is to solve the deficiencies existing in the prior art, and to propose a method for automatically identifying abnormal refueling events based on a refueling scenario.

[0008] To achieve the above purpose, the present invention adopts the following technical solutions: The working content of a method for automatically identifying abnormal refueling events based on a refueling scenario specifically includes the following steps: S1: Collect data through the sensor network and the communication module; The sensor network includes a fuel gun sensor, a fuel tank liquid level sensor, a GPS module, an in-vehicle terminal device; The fuel gun sensor is used to record the refueling start time t1, the end time t2, and the refueling volume M. The duration of the refueling is t2 - t1, with the unit of seconds, and the unit of the refueling volume M is liters; The fuel tank liquid level sensor is an ultrasonic fuel tank liquid level sensor, which measures the oil level L in the fuel tank in real time. The unit of the liquid level L is centimeters; The GPS module is used to record the vehicle position; The in-vehicle terminal device: integrates an accelerometer and a fuel consumption calculation module for calculating the driving state and fuel consumption; The driving state includes the driving mileage S, vehicle speed V, and rotational speed G; The communication module is an OBD-II interface for transmitting the data collected by the sensor network, GPS module, and in-vehicle terminal device to the cloud through 5G communication. The OBD-II interface is electrically connected to the gas station method to collect the transaction record data of the gas station; The transaction record data includes the refueling time T, refueling volume N, and refueling personnel information ID; The data includes the data collected by the sensor network and the transaction record data; S2: Upload the collected data to the in-vehicle terminal device through the communication module; The data is uploaded to the in-vehicle terminal device through the OBD-II interface; The OBD-II interface uses the MQTT protocol; The in-vehicle terminal device deploys edge nodes; S3: The in-vehicle terminal device cleans the received data and eliminates invalid information; The data includes the speed V, fuel level L, and vehicle position; The cleaning includes filtering invalid information and smoothing the level information; The invalid information: the level value exceeds the maximum level of the fuel tank or the GPS signal is lost; The smoothing of the level information: calculates the average value using a sliding window, and the window size is 5 seconds. The specific content is as follows: L-smooth(t) =

L(t-2) + L(t-1) + L(t) + L(t+1) + L(t+2)

[0009] S4: The in-vehicle terminal device determines and divides the parking state of the vehicle according to the cleaned data; Includes the following steps: S41: Determine the parking state of the vehicle; If the speed V = 0 and the ACC state is "on" or "off" for more than 30 seconds, it is marked as the parking state; S42: Divide the parking interval of the vehicle; Specifically includes the following steps: S421: Define the parking interval of the vehicle: The endpoints of the parking interval are defined as the starting point and the ending point; The starting point: the moment when the speed changes from non-zero to zero; the ending point: the moment when the speed changes from zero to non-zero; S422: Extract the information of the fuel level L within the interval; Record the fuel level L_before[T_start - 300s, T_start] five minutes before parking; Record the fuel level L_parking[T_start, T_end] during parking; Record the fuel level L_after[T_end, T_end + 300s] five minutes after parking; S5: Detect the change in the fuel level in the vehicle fuel tank; Specifically, it includes the following steps: S51: Calculate the change amount of the fuel level before and after parking; Take the average value of the fuel level within the last minute before parking: L_pre = mean(L_before[-60s, T_start]); Take the average value of the fuel level within the first minute after parking:: L_post = mean(L_after[T_end, T_end + 60s]); The change amount of the fuel level before and after parking △L = L_post - L_pre; S52: Set the threshold for the change amount of the fuel level; The threshold includes a refueling threshold Th_fuel and a fuel theft threshold Th_steal; Refueling threshold Th_fuel: The minimum change amount △L1 of the sudden rise in the liquid level; Fuel theft threshold Th_steal: The minimum change amount △L2 of the sudden drop in the liquid level; The threshold can be dynamically adjusted according to the fuel tank capacity and sensor accuracy; S53: Determine three events of fuel theft, refueling, and normal according to the threshold; If △L ≥ △L1, it is marked as a refueling event; If △L ≤ △L2, it is marked as a fuel theft event; If △L1 ≤ △L ≤ △L2, it is marked as a normal event; S54: During the parking interval, find the starting point T_change of the liquid level change; Calculate the first-order difference for the parking interval: ΔL_t = L(t) - L(t - 1); If ΔL_t > Th_fuel / Δt, then T_change is the refueling start time; if ΔL_t < Th_steal / Δt, then T_change is the fuel theft start time; The Δt is the sampling interval time, usually 1 second; Event duration: from T_change to the moment T_end_change when the liquid level stabilizes.

[0010] S6: Verify and output the three events of oil theft, refueling, and normal; Specifically, it includes the following steps: S61: The in-vehicle terminal device extracts the vehicle location information from document W; S62: The in-vehicle terminal device verifies the refueling event; When the in-vehicle terminal device determines that a refueling event has occurred, it checks whether the GPS location is within the range of a known gas station. The range is 500 meters. If the GPS location is within the range of the known gas station, it is determined as refueling; if the GPS location is not within the range of the known gas station, the in-vehicle terminal device sends an abnormal alarm message to the user terminal platform; S63: The in-vehicle terminal device verifies the oil theft event; When the in-vehicle terminal device determines that an oil theft event has occurred, it compares whether the GPS location is in the gas station area. If the GPS location is not in the gas station area, it is determined as an oil theft event and an oil theft alarm message is sent to the user terminal platform; if the GPS location is in the gas station area, the in-vehicle terminal device sends an abnormal alarm message to the user terminal platform; S64: The in-vehicle terminal device sends the output result to the user terminal platform; The user terminal platform: provides a visual fleet management interface, including refueling event records, abnormal event alarms, and data statistical analysis.

[0011] The output result includes the event type, start time T_change, end time T_end_change, fuel quantity and liquid level change amount ΔL, and vehicle location.

[0012] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention provides a method for identifying abnormal events based on the refueling scenario. Through the deep combination of hardware and algorithms, relying on the Internet of Things, sensor technology, and edge computing, the real-time monitoring of the entire refueling process is realized, and the real-time information of the entire refueling process is sent to the user terminal platform in a timely manner.

[0013] The method proposed by the present invention; the present invention deploys a high-precision sensor network in vehicles and gas stations, including fuel gun sensors, fuel tank liquid level sensors, GPS modules, and in-vehicle terminal devices. Through the collaborative work of these hardware components, the multi-source and accuracy of refueling data are ensured.

[0014] ​The method proposed by the present invention constructs a closed-loop refueling management system through real-time hardware perception, intelligent algorithm analysis, and efficient user-side feedback. It can not only detect abnormalities in a timely manner but also provide decision-making support for fleet managers, achieving cost optimization and efficiency improvement. Description of the Drawings

[0015] Figure 1 It is a flowchart of the working steps of a method for automatically identifying abnormal refueling events based on a refueling scenario according to the present invention. Detailed Embodiments

[0016] To further understand the purpose, structure, features, and functions of the present invention, the following is a detailed description in conjunction with embodiments.

[0017] As Figure 1 shown, the working content of a method for automatically identifying abnormal refueling events based on a refueling scenario specifically includes the following steps: S1: Collect data through the sensor network and the communication module; The sensor network includes a fuel gun sensor, a fuel tank liquid level sensor, a GPS module, in-vehicle terminal device; The fuel gun sensor is used to record the refueling start time t1, the end time t2, and the refueling quantity M. The duration of refueling is t2 - t1, in seconds, and the unit of the refueling quantity M is liters; The fuel tank liquid level sensor is an ultrasonic fuel tank liquid level sensor that measures the oil quantity liquid level L in the fuel tank in real time. The unit of the liquid level L is centimeters; By calculating the change value of the oil quantity liquid level L, it can be calculated whether the refueling quantity M per time is equal to the oil quantity received in the fuel tank; ensure that each time refueling occurs, the oil added by the fuel gun to the fuel tank corresponds to the oil quantity liquid level L information collected by the ultrasonic fuel tank liquid level sensor.

[0018] The GPS module is used to record the vehicle position; The in-vehicle terminal device: integrates an accelerometer and a fuel consumption calculation module, and is used to calculate the driving state and fuel consumption; The driving state includes the driving mileage S, the vehicle speed V, and the rotation speed G; The communication module is an OBD-II interface, which is used to transmit the data collected by the sensor network, the GPS module, and the in-vehicle terminal device to the cloud through 5G communication. The OBD-II interface is electrically connected to the gas station method to collect the transaction record data of the gas station; The OBD-II interface adopts the MQTT protocol, supports data transmission in low-bandwidth environments, with a message loss rate lower than 0.1%, and is suitable for cross-regional management of vehicle fleets.

[0019] The transaction record data includes refueling time T, refueling volume N, and refueling personnel information ID; The data includes data collected by the sensor network and transaction record data; S2: Upload the collected data to the in-vehicle terminal device through the communication module; The data is uploaded to the in-vehicle terminal device through the OBD-II interface; The OBD-II interface adopts the MQTT protocol; The in-vehicle terminal device deploys edge nodes. Edge nodes are deployed at the in-vehicle terminal to preprocess sensor data and perform preliminary anomaly detection, reducing the computing pressure on the cloud and controlling the end-to-end delay within 500 ms; S3: The in-vehicle terminal device cleans the received data and eliminates invalid information; The data includes speed V, fuel level L, and vehicle position; Clean the extracted data, eliminate invalid information, and obtain valid information; The cleaning includes filtering invalid information and smoothing the level information; The invalid information: the level value exceeds the maximum level of the fuel tank or the GPS signal is lost; at this time, the in-vehicle terminal device sends an alarm message to the user terminal platform, and the user terminal platform shows an abnormal data prompt.

[0020] The smoothed level information: use a sliding window to calculate the average value to reduce the influence of sensor noise. The window size is 5 seconds, and the specific content is as follows: L-smooth(t) =

L(t-2) + L(t-1) + L(t) + L(t+1) + L(t+2)

[0021] S4: The in-vehicle terminal device determines the parking state of the vehicle and divides the intervals based on the cleaned data; It includes the following steps: S41: Determine the parking state of the vehicle; If the speed V = 0 and the ACC status is "on" or "off" for more than 30 seconds, it is marked as the parking state; S42: Divide the parking intervals of the vehicle; Specifically, it includes the following steps: S421: Define the parking intervals of the vehicle: The endpoints of the parking intervals are defined as the starting point and the ending point; The starting point: the moment when the speed changes from non-zero to zero; the ending point: the moment when the speed changes from zero to non-zero; S422: Extract the information of the fuel level L within the interval; Record the fuel level L_before [T_start - 300s, T_start] five minutes before parking; Record the fuel level L_parking [T_start, T_end] during parking; Record the fuel level L_after [T_end, T_end + 300s] five minutes after parking; S5: Detect the change in the fuel level within the vehicle fuel tank; Specifically, it includes the following steps: S51: Calculate the change amount of the fuel level before and after parking; Take the average value of the fuel level within the last minute before parking: L_pre = mean(L_before[-60s, T_start]); Take the average value of the fuel level within the first minute after parking:: L_post = mean(L_after[T_end, T_end + 60s]); Take the average value of the fuel level within one minute before and after parking to reduce the influence of other factors on the fuel level; The change amount of the fuel level before and after parking △L = L_post - L_pre; S52: Set the threshold for the change amount of the fuel level; The threshold includes a refueling threshold Th_fuel and an oil theft threshold Th_steal; Refueling threshold Th_fuel: the minimum change amount △L1 of the sudden rise in the liquid level; Oil theft threshold Th_steal: the minimum change amount △L2 of the sudden drop in the liquid level; The threshold can be dynamically adjusted according to the fuel tank capacity and sensor accuracy; By setting the threshold, the in-vehicle terminal device can determine three events: oil theft, refueling, and normal; S53: Determine the three events of oil theft, refueling, and normal according to the threshold; If △L ≥ △L1, it is marked as a refueling event; If △L ≤ △L2, it is marked as an oil theft event; If △L1 ≤ △L ≤ △L2, it is marked as a normal event; S54: During the parking interval, find the starting point T_change of the liquid level change; Calculate the first-order difference for the parking interval: ΔL_t = L(t) - L(t-1); If ΔL_t > Th_fuel / Δt, then T_change is the start time of refueling; if ΔL_t < Th_steal / Δt, then T_change is the start time of fuel theft; The Δt is the sampling interval time, usually 1 second; Determine the start time of refueling or fuel theft in this way; Event duration: from T_change to the moment T_end_change when the liquid level is stable.

[0022] S6: Verify and output the three events of fuel theft, refueling, and normal; Specifically, it includes the following steps: S61: The in-vehicle terminal device extracts the vehicle position information from document W; S62: The in-vehicle terminal device verifies the refueling event; When the in-vehicle terminal device determines that a refueling event has occurred, check whether the GPS position is within the range of a known gas station, and the range is 500 meters. If the GPS position is within the range of the known gas station, it is determined as refueling; if the GPS position is not within the range of the known gas station, the in-vehicle terminal device sends an abnormal alarm message to the user terminal platform; S63: The in-vehicle terminal device verifies the fuel theft event; When the in-vehicle terminal device determines that a fuel theft event has occurred, compare whether the GPS position is in the gas station area. If the GPS position is not in the gas station area, it is determined as a fuel theft event, and a fuel theft alarm message is sent to the user terminal platform; if the GPS position is in the gas station area, the in-vehicle terminal device sends an abnormal alarm message to the user terminal platform; S64: The in-vehicle terminal device sends the output result to the user terminal platform; The user terminal platform: provides a visual fleet management interface, including refueling event records Abnormal event alarms and data statistical analysis.

[0023] The output result includes the event type, start time T_change, end time T_end_change, fuel quantity and liquid level change amount ΔL, and vehicle position.

[0024] The present invention has been described by the above related embodiments. However, the above embodiments are only examples for implementing the present invention. It must be pointed out that the disclosed embodiments do not limit the scope of the present invention. On the contrary, modifications and refinements made without departing from the spirit and scope of the present invention fall within the scope of patent protection of the present invention.

Claims

1. A method for automatically identifying abnormal refueling events based on a refueling scenario, characterized in that: Specifically, it includes the following steps: S1: Collect data through the sensor network and the communication module; S2: Upload the collected data to the vehicle terminal device through the communication module; S3: The vehicle terminal device cleans the received data and eliminates invalid information; S4: The vehicle terminal device determines the parking state and divides the interval of the vehicle according to the cleaned data; S5: Detect the change in the fuel level in the vehicle fuel tank; Specifically, it includes the following steps: S51: Calculate the change amount of the fuel level before and after parking; Take the average value of the fuel level in the last minute before parking: L_pre = mean(L_before[-60s, T_start]); Take the average value of the fuel level in the first minute after parking: L_post = mean(L_after[T_end, T_end+60s]); The change amount of the fuel level before and after parking △L = L_post - L_pre; S52: Set the threshold of the change amount of the fuel level; The threshold includes the refueling threshold Th_fuel and the fuel theft threshold Th_steal; Refueling threshold Th_fuel: The minimum change amount △L1 of the sudden rise of the liquid level; Fuel theft threshold Th_steal: The minimum change amount △L2 of the sudden drop of the liquid level; The threshold can be dynamically adjusted according to the fuel tank capacity and the sensor accuracy; S53: Determine the three events of fuel theft, refueling, and normal according to the threshold; If △L≥△L1, it is marked as a refueling event; If △L≤△L2, it is marked as a fuel theft event; If △L1≤△L≤△L2, it is marked as a normal event; S54: In the parking interval, find the starting point T_change of the liquid level change; Calculate the first-order difference for the parking interval: ΔL_t = L(t) - L(t-1); If ΔL_t > Th_fuel / Δt, then T_change is the start time of refueling; if ΔL_t < Th_steal / Δt, then T_change is the start time of fuel theft; The Δt is the sampling interval time, usually 1 second; Event duration: From T_change to the moment T_end_change when the liquid level is stable; S6: Verify and output the three events of fuel theft, refueling, and normal; Specifically, it includes the following steps: S61: The vehicle terminal device extracts the vehicle position information from the document W; S62: The vehicle terminal device verifies the refueling event; S63: The vehicle terminal device verifies the fuel theft event; S64: The vehicle terminal device sends the output result to the user terminal platform.

2. The method for automatically identifying abnormal refueling events based on a refueling scenario according to claim 1, wherein: The specific content of step S3 includes: The cleaning includes filtering invalid information and smoothing the liquid level information; The invalid information: The liquid level value exceeds the maximum liquid level of the fuel tank or the GPS signal is lost; The smoothing of the liquid level information: Calculate the average value using a sliding window, and the window size is 5 seconds. The specific content is as follows: L-smooth(t) = 【L(t-2) + L(t-1) + L(t) + L(t+1) + L(t+2)】 / 5.

3. The method for automatically identifying abnormal refueling events based on a refueling scenario according to claim 1, wherein: Step S4 Specifically, it includes the following steps: S41: Determine the parking state of the vehicle; S42: Divide the parking section of the vehicle; Step S42 specifically includes the following steps: S421: Define the parking section of the vehicle: S422: Extract the information of the fuel level L within the section; Record the fuel level L_before [T_start - 300s, T_start] five minutes before parking; Record the fuel level L_parking [T_start, T_end] during parking; Record the fuel level L_after [T_end, T_end + 300s] five minutes after parking.

4. The method for automatically identifying abnormal refueling events based on a refueling scenario according to claim 3, wherein: Step S422 specifically includes the following steps: Record the fuel level L_before [T_start - 300s, T_start] five minutes before parking; Record the fuel level L_parking [T_start, T_end] during parking; Record the fuel level L_after [T_end, T_end + 300s] five minutes after parking.

5. The method for automatically identifying abnormal refueling events based on a refueling scenario according to claim 1, wherein: Step S62 specifically includes the following steps: When the in-vehicle terminal device determines that a refueling event has occurred, check whether the GPS position is within the range of a known gas station, and the range is 500 meters. If the GPS position is within the range of the known gas station, it is determined that refueling has occurred; if the GPS position is not within the range of the known gas station, the in-vehicle terminal device sends an abnormal alarm message to the user terminal platform.

6. The method for automatically identifying abnormal refueling events based on a refueling scenario according to claim 1, wherein: Step S63 specifically includes the following steps: When the in-vehicle terminal device determines that an oil theft event has occurred, compare whether the GPS position is in the gas station area. If the GPS position is not in the gas station area, it is determined that an oil theft event has occurred, and an oil theft alarm message is sent to the user terminal platform; if the GPS position is in the gas station area, the in-vehicle terminal device sends an abnormal alarm message to the user terminal platform.

Images