Intelligent vehicle scheduling optimization system based on big data

Abstract

The invention relates to the technical field of vehicle scheduling optimization, and discloses an intelligent vehicle scheduling optimization system based on big data, which comprises a multi-source data acquisition unit, a fuel consumption increment calculation unit, a fuel charge optimization unit and an information feedback module, the multi-source data acquisition unit is used for acquiring oil consumption state, real-time traffic, weather and road geographic data of truck driving; and the oil consumption increment calculation unit is used for dynamically predicting extra oil consumption caused by real-time traffic, weather and road geographic data factors based on a machine learning model. According to the method, the path distance of the truck and fuel consumption increase caused by extra weather change, traffic jam, road maintenance and red light waiting time in the path distance are obtained according to the big data, the refueling amount of each time is planned according to the fuel consumption increase, and the situation that the fuel consumption of the truck is increased due to excessive remaining amount of a fuel tank of the truck is avoided by planning the proper refueling amount. And oil is stolen when a user has a rest in a service area.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle dispatch optimization, and in particular to an intelligent vehicle dispatch optimization system based on big data. Background Art

[0002] During long-distance transportation, large trucks often have too much fuel in their tanks due to empirical refueling, which makes them vulnerable to oil theft when parked at service areas (according to statistics, my country's annual losses from oil theft to trucks exceed 1 billion yuan). Traditional fuel consumption estimates rely on static road condition data and do not dynamically integrate real-time variables such as weather, congestion, and road slope, resulting in a high error rate in refueling planning. Fixed refueling strategies cannot adapt to sudden road conditions (such as temporary road closures and extreme weather), which may lead to the risk of running out of fuel midway or having too much fuel remaining.

[0003] The existing patent discloses a large vehicle anti-theft method based on artificial intelligence video recognition (publication number CN111046822A), including: a). Video of the thief's posture in various weather environments such as service areas, loading points, and unloading points where large vehicles are parked, and capture images; b). Manually annotate the image, and the annotated points are people who appear near the location of large vehicles; c). First, establish a training model, train according to the established neural network model, and obtain a special target detection model for detecting whether there are suspects approaching large vehicles; d). Test the recognition accuracy and improve and optimize. The technology disclosed in this patent cannot completely avoid the problem of oil theft, especially in remote service areas with sparse traffic at night. Due to the poor quality of network signals in remote areas, the real-time monitoring and analysis of the network model is not timely, and the vehicle tank will still be damaged. Summary of the invention

[0004] The main technical problem solved by the present invention is to provide an intelligent vehicle scheduling optimization system based on big data, which solves the problems in the above-mentioned background technology.

[0005] To solve the above technical problems, according to one aspect of the present invention, more specifically, an intelligent vehicle dispatch optimization system based on big data includes a multi-source data acquisition unit, a fuel consumption increment calculation unit, a refueling quantity optimization unit, and an information feedback module; The multi-source data acquisition unit is used to obtain the fuel consumption status, real-time traffic, weather and road geographic data of the truck; The fuel consumption increment calculation unit dynamically predicts the additional fuel consumption caused by real-time traffic, weather and road geographic data factors based on the machine learning model; The refueling quantity optimization unit is used to generate a refueling quantity and service area recommendation strategy.

[0006] Further, the intelligent vehicle scheduling optimization system further includes an oil theft risk estimation module and a historical data collection module.

[0007] Further, the oil theft risk estimation module is used to judge the risk of oil theft for a truck based on the remaining amount of gasoline or diesel in the truck's fuel tank, the frequency of oil theft for trucks in the current 100-kilometer area, and the night traffic flow in the area where the truck is currently parked. Then, we have: ; In the formula, represents the risk coefficient of oil theft when the truck is parked in the current area, represents the remaining fuel volume in the truck's current fuel tank, represents the frequency of oil theft for trucks in the current 100-kilometer area, represents the night traffic flow in the area where the truck is currently parked.

[0008] Further, when , it indicates that the risk of oil theft for the truck parked in the current area is relatively high; When , it indicates that there is a risk of oil theft for the truck parked in the current area; When , it indicates that the risk of oil theft for the truck parked in the current area is very low.

[0009] Further, the historical data collection module is used to collect the information of the dates, locations and service areas where gasoline or diesel of national trucks was stolen stored in the cloud database and the traffic flow information in the area where the truck is currently parked.

[0010] Further, the fuel consumption increment calculation unit is used to calculate the additional fuel consumption per kilometer of the truck according to the influence of road traffic congestion degree, road average gradient, red light waiting time and wind speed. Then, we have: ; In the formula, represents the additional fuel consumption per kilometer of the truck; , , , respectively represent the weight coefficients of the influence of road traffic congestion, road average gradient, red light waiting time and wind speed; represents the traffic congestion coefficient; represents the road average gradient; represents the density of red lights in the road section; represents the headwind speed in the vehicle driving direction.

[0011] Further, the refueling quantity optimization unit calculates the required fuel quantity for the remaining mileage according to the feedback of the fuel consumption increment calculation unit, and we have: ; In the formula, Indicates the recommended refueling amount of the refueling amount optimization unit; Indicates the theoretical fuel consumption required to complete the remaining journey; Indicates the amount of safety redundant fuel, used to cope with unexpected road conditions; Indicates the current remaining fuel tank amount, which is obtained in real time through the vehicle fuel level sensor; then: ; In the formula, Indicates the remaining distance from the current location to the destination or the next recommended service area; Indicates the base fuel consumption when the vehicle is empty or fully loaded.

[0012] Furthermore, the information feedback module is used to send the truck driver the next refueling amount and relevant information of the next refueling service area based on the refueling information analyzed by the refueling amount optimization unit.

[0013] The present invention provides an intelligent vehicle dispatch optimization system based on big data. Compared with the prior art, the present method has the following effects: 1. The present invention obtains the route of large trucks and the increase in fuel consumption caused by additional weather changes, traffic jams, road maintenance and red light waiting time during the journey based on big data, and plans the amount of refueling each time according to the increase in fuel consumption. By planning the appropriate amount of refueling, the problem of excessive remaining fuel in the fuel tank of large trucks and theft of fuel when resting in service areas can be avoided.

[0014] 2. The present invention relates refueling volume planning to the probability of oil theft, and introduces public security data to dynamically analyze the risk of oil theft by trucks in the current area. The risk factor can further assist drivers in trip scheduling and fuel tank oil volume planning.

[0015] 3. The present invention quantifies the risk factor based on the remaining amount in the fuel tank, the frequency of oil theft in the area and the traffic volume at night, and then assists the driver in choosing a low-risk service area, which can greatly avoid the economic loss of the truck driver.

[0016] 4. The present invention obtains real-time traffic, weather, and road geographic information through multi-source data, and accurately predicts additional fuel consumption through machine learning models, thereby improving the accuracy of refueling quantity planning. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a flow chart of the vehicle dispatch optimization system in the present invention; Figure 2 A relationship diagram between the oil theft risk factor and the remaining amount in the oil tank in the present invention; Figure 3Graph showing the relationship between additional fuel consumption and traffic congestion coefficient in the present invention; Figure 4 Graph showing the relationship between the recommended fuel filling amount and the remaining mileage in the present invention. Detailed implementation manners

[0018] To make the technical solution of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] Embodiment 1 As Figure 1 shown, according to one aspect of the present invention, an intelligent vehicle scheduling optimization system based on big data is provided, including a multi-source data acquisition unit, a fuel consumption increment calculation unit, a fuel filling amount optimization unit, and an information feedback module; the multi-source data acquisition unit is used to obtain the fuel consumption status, real-time traffic, weather, and road geography data of the truck driving; the fuel consumption increment calculation unit is used to dynamically predict the additional fuel consumption caused by real-time traffic, weather, and road geography data factors based on a machine learning model; the fuel filling amount optimization unit is used to generate a fuel filling amount and service area recommendation strategy.

[0020] Embodiment 2 As Figure 1 、 2 shown, the intelligent vehicle scheduling optimization system further includes an oil theft risk estimation module and a historical data acquisition module; the historical data acquisition module is used to collect the national truck gasoline or diesel theft dates, theft locations, and service area information stored in the cloud database and the traffic flow information of the current truck parking area. The oil theft risk estimation module is used to judge the risk of the truck being stolen according to the remaining amount of gasoline or diesel in the truck fuel tank, the frequency of the truck being stolen in the current 100-kilometer area, and the night traffic flow of the current truck parking area. Then there is: ; In the formula, represents the risk coefficient of the truck being stolen in the current area, represents the remaining fuel amount in the current truck fuel tank, represents the frequency of the truck being stolen in the current 100-kilometer area, represents the night traffic flow of the current truck parking area.

[0021] Among them, the risk coefficient of the truck being stolen in the current area is calculated for any time. The remaining fuel amount in the current truck fuel tank takes , the frequency of the truck being stolen in the current 100-kilometer area takes (times / month), and the night traffic flow of the current truck parking area takes (vehicles / hour. The night traffic flow is calculated from 22:00 to 6:00 the next day). Then there is:

[0022] It can be known from the above calculations that the risk coefficient of the truck parked in the current area being stolen of fuel is . And by comparing multiple sets of data, there are: Table 1 Partial implementation parameters and risk status of the truck being stolen of fuel It can be known from the data in Table 1 above that when the sample data tends to infinity, there will be a dividing line for dividing whether the truck has the risk of being stolen of fuel by the risk coefficient P, that is, when , it means that the risk of the truck parked in the current area being stolen of fuel is relatively high; when , it means that there is a risk of the truck parked in the current area being stolen of fuel; when , it means that the risk of the truck parked in the current area being stolen of fuel is very small.

[0023] Example 3 As Figure 3 shown, the fuel consumption increment calculation unit is used to calculate the additional fuel consumption per kilometer of the truck according to the influence of road traffic congestion degree, road average slope, red light waiting time, and wind speed. Then there is: ; In the formula, represents the additional fuel consumption per kilometer of the truck; , , , respectively represent the weight coefficients of the influence of road traffic congestion, road average slope, red light waiting time, and wind speed. The calculation methods of the fuel consumption influence weight coefficients of each factor are as follows: 1), Training the historical driving data through machine learning.

[0024] 2), The training data is based on the actual fuel consumption records of the vehicle on 100 different congested, sloped, and red light sections.

[0025] 3), Output of the data. The trained empirical data are: (liters / km·congestion coefficient), (liters / km·% slope), (liters / km·red light density), (liters / km·m / s wind speed).

[0026] represents the traffic congestion coefficient (the traffic congestion coefficient, with a value range of 0-1, where 0 means smooth traffic and 1 means severe congestion. The traffic congestion coefficient can obtain the congestion index through the real-time traffic API (such as Gaode Map)).

[0027] Indicates the average slope of the road (the percentage of the height difference of the road section climbing to the horizontal distance is calculated through GIS elevation data. For example, if the climbing is 10 meters / kilometer, the slope is 1%); Indicates the density of traffic lights in a road section (the density of traffic lights is calculated based on the number of traffic lights per kilometer according to map data. For example, if there are 8 traffic lights within 5 kilometers of a road section, the density is 1.6 per kilometer); Indicates the headwind speed in the direction of vehicle travel (the headwind speed is obtained through the weather API in real time).

[0028] like Figure 4 As shown, the refueling amount optimization unit calculates the required fuel amount for the remaining mileage based on the feedback from the fuel consumption increment calculation unit, as follows: ; In the formula, Indicates the recommended refueling amount of the refueling amount optimization unit; Indicates the theoretical fuel consumption required to complete the remaining journey; Indicates the amount of safe redundant fuel, used to cope with unexpected road conditions (the default safety factor is 20% redundant. It can also be adjusted dynamically based on the security risk of the service area: only 10% redundant in high-risk areas to reduce the amount of stolen fuel, and 30% redundant in low-risk areas to improve fault tolerance); Indicates the current remaining fuel tank amount, which is obtained in real time through the vehicle fuel level sensor; then: ; In the formula, Indicates the remaining distance from the current location to the destination or the next recommended service area; Indicates the base fuel consumption when the vehicle is empty or fully loaded.

[0029] The remaining fuel volume in a truck tank (liters), it takes D = 300 kilometers to reach the destination, and the recommended fuel consumption of the fuel optimization unit is calculated. (L / km). The real-time environmental data for this journey include: Traffic congestion coefficient , (indicates congestion), the average road slope is , , the density of traffic lights in the road section is , , the wind speed against the vehicle's direction of travel is , , then we have:

[0030] From the above calculation, we can know that the additional fuel consumption per kilometer of the truck is (L / km). Calculate the theoretical fuel consumption required for the truck to complete the remaining journey, and we have:

[0031] Then it can be known that the theoretical fuel consumption required for the truck to complete the remaining journey is (L). Then calculate the recommended fuel filling amount of the fuel filling amount optimization unit, and we have:

[0032] From the above calculations, it can be known that the recommended fuel filling amount of the fuel filling amount optimization unit is (L).

[0033] Embodiment 4 An information feedback module, configured to send the relevant information of the next fuel filling amount and the next fuel filling service area to the truck driver according to the fuel filling information analyzed by the fuel filling amount optimization unit.

[0034] The above-described embodiments merely represent several implementation manners of the present invention. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the present invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention patent shall be subject to the appended claims.

Claims

1. An intelligent vehicle dispatch optimization system based on big data, characterized in that: It includes a multi-source data acquisition unit, a fuel consumption increment calculation unit, a refueling quantity optimization unit and an information feedback module; The multi-source data acquisition unit is used to obtain the fuel consumption status, real-time traffic, weather and road geographic data of the truck; The fuel consumption increment calculation unit dynamically predicts the additional fuel consumption caused by real-time traffic, weather and road geographic data factors based on the machine learning model; The refueling quantity optimization unit is used to generate a refueling quantity and service area recommendation strategy.

2. The intelligent vehicle dispatch optimization system based on big data according to claim 1 is characterized by: The intelligent vehicle dispatch optimization system also includes an oil theft risk estimation module and a historical data collection module.

3. The intelligent vehicle dispatch optimization system based on big data according to claim 2 is characterized by: The oil theft risk estimation module is used to determine the risk of oil theft on a truck based on the remaining gasoline or diesel in the truck tank, the frequency of oil theft on the truck in the current 100-kilometer area, and the nighttime traffic volume in the truck's current parking area. The following are the risks: ； In the formula, Indicates the risk factor of oil being stolen when the truck is parked in the current area. Indicates the current remaining fuel volume in the truck tank. Indicates the frequency of oil theft on trucks in the current 100-kilometer area. Indicates the nighttime traffic volume in the area where the truck is currently parked.

4. The intelligent vehicle dispatch optimization system based on big data according to claim 3 is characterized by: when When , it means that the truck is parked in the current area and the risk of oil being stolen is high; when When , it means that there is a risk of oil being stolen in the current area where the truck is parked; when When , it means that the risk of oil being stolen when the truck is parked in the current area is very small.

5. The intelligent vehicle dispatch optimization system based on big data according to claim 2 is characterized in that: The historical data collection module is used to collect the national truck gasoline or diesel theft date, theft location and service area information and the traffic flow information of the truck's current parking area stored in the cloud database.

6. The intelligent vehicle dispatch optimization system based on big data according to claim 1 is characterized by: The fuel consumption increment calculation unit is used to calculate the additional fuel consumption per kilometer of the truck according to the degree of road traffic congestion, the average road slope, the waiting time at the traffic light and the influence of wind speed, so: ； In the formula, It indicates the extra fuel consumption per kilometer of the truck; , , , They represent the weight coefficients of road traffic congestion, average road slope, waiting time at traffic lights, and wind speed respectively; represents the traffic congestion coefficient; Indicates the average slope of the road; Indicates the density of traffic lights in the road section; Indicates the wind speed against the direction of vehicle travel.

7. The intelligent vehicle dispatch optimization system based on big data according to claim 6 is characterized by: The refueling amount optimization unit calculates the required fuel amount for the remaining mileage according to the feedback from the fuel consumption increment calculation unit, as follows: ； In the formula, Indicates the recommended refueling amount of the refueling amount optimization unit; Indicates the theoretical fuel consumption required to complete the remaining journey; Indicates the amount of safety redundant fuel, used to cope with unexpected road conditions; Indicates the current remaining fuel tank amount, which is obtained in real time through the vehicle fuel level sensor; then: ； In the formula, Indicates the remaining distance from the current location to the destination or the next recommended service area; Indicates the base fuel consumption when the vehicle is empty or fully loaded.

8. The intelligent vehicle dispatch optimization system based on big data according to claim 1 is characterized by: The information feedback module is used to send the next refueling amount and related information of the next refueling service area to the truck driver according to the refueling information analyzed by the refueling amount optimization unit.

Images