Road vehicle passing management method based on data prediction

Abstract

The invention discloses a road vehicle passing management method based on data prediction, and the method comprises the following steps: S1, giving a starting point and a destination of a vehicle, and a plurality of routes from the starting point to the destination, and each route comprises a plurality of roads; s2, collecting historical traffic flow information, historical weather information and historical date information of each section of road in each route, and time information used for passing the section of road, and constructing a prediction model of each section of road based on the collected information and training the prediction model; and S3, predicting traffic flow information by using the trained prediction model, and performing route selection according to time information used by a road to realize traffic management of vehicles. According to the method, the prediction model of each section of road can be constructed for each route and trained, the prediction model on each section of road can be obtained to predict the traffic flow and the time required for passing through the road section, and the optimal route is selected in a mode of weighting the traffic flow, the passing time and the mileage.

Description

TECHNICAL FIELD

[0001] The present application relates to vehicle passage association, in particular to a road vehicle passage management method based on data prediction. BACKGROUND

[0002] With the continuous progress and development of modern science and technology, the operation vehicles are increasing, which brings many conveniences for people's travel and transportation. The present starting point and destination of the fixed operation vehicles are generally divided into the following two kinds: the first kind is the fixed operation vehicle of route, which does not allow real-time route adjustment; the second kind is the operation vehicle of optional route, such as the designated starting point and the key point of the truck, etc. For the operation vehicle of optional route, the route needs to be selected according to the actual situation to achieve the balance of total mileage, total time and congestion degree, but the congestion degree and the passing time of each section of the route are variable, which leads to the difficulty in optimal selection during planning, and does not utilize the communication management of road vehicles. SUMMARY

[0003] The present application aims at overcoming the shortcomings of the prior art, and provides a road vehicle passage management method based on data prediction, which can obtain the predicted traffic flow and the time required for passing each section of road by constructing and training the prediction model of each section of road for each route, and select the optimal route by weighting the traffic flow, the passing time and the mileage.

[0004] The purpose of the present application is achieved by the following technical scheme: a road vehicle passage management method based on data prediction, comprising the following steps: S1. Given the vehicle starting point and destination, and multiple routes from the starting point to the destination, each route contains multiple sections of road; S2. Collect the historical traffic flow information, historical weather information, historical date information and time information of each section of road in each route, and construct and train the prediction model of each section of road based on the collected information; S3. For each section of road in each route, the trained prediction model is used to predict the traffic flow information and the time information of passing the road, and the route selection is performed by comprehensively considering the route length, the traffic flow information of each section of road in the route and the time of passing the section of road, so as to realize the passage management of the vehicle.

[0005] The beneficial effects of the present application are: the present application can predict the departure time of the starting point of each road section in the route by constructing the prediction model of each road section of each route and training, so as to obtain the information of the road section, so as to more accurately estimate the traffic flow and passing time of each road section, and finally select the optimal route by weighting the traffic flow, passing time and mileage. BRIEF DESCRIPTION OF DRAWINGS

[0006] Figure 1 The method flowchart of the present application. DETAILED DESCRIPTION

[0007] The technical solutions of the present application will be described in detail below in combination with the drawings, but the protection scope of the present application is not limited to the following description.

[0008] As shown in the figure, a road vehicle passing management method based on data prediction comprises the following steps: Figure 1 S1. Given the vehicle departure point and destination, and a plurality of routes from the departure point to the destination, each route comprising a plurality of road sections; S2. Collecting historical traffic flow information, historical weather information, historical date information of each road section in each route, and time information for passing through the road section, and constructing a prediction model for each road section based on the collected information and training; The step S2 comprises: S201. For any road section, collect historical traffic flow information, time information for passing through the road section, historical weather information and historical date information within a period T, and construct a training sample for the current period; Let the period contain N time periods, then: The historical traffic flow information is denoted as , n=1,2,…,N; , n=1,2,…,N; , n=1,2,…,N; wherein , n=1,2,…,N; wherein , n=1,2,…,N; wherein , n=1,2,…,N; wherein , n=1,2,…,N; wherein , n=1,2,…,N; wherein , n=1,2,…,N; wherein , n=1,2,…,N; wherein , n=1,2,…,N; wherein indicates weather information in the nth time period, indicates overcast, indicates sunny or cloudy, indicates rain or snow, indicates extreme weather conditions, including hail, tornado and thunderstorm; The historical date information is recorded as , n = 1, 2, …, N; wherein indicates the historical date information in the nth time period, indicates a working day, indicates a non-working day; N training samples of the current period are constructed, wherein the nth time period training sample is recorded as , n = 1, 2, …, N, wherein, is a sample feature vector, is a sample label vector; S202. Repeat step S201 to obtain training samples in multiple periods T and add them to the training sample set of the current road segment; S203. Based on the deep neural network algorithm, a prediction model of the current road segment is constructed, and the training sample set of the current road segment is used to train the prediction model to obtain a trained prediction model; S204. For each road segment in each route, steps S201-S203 are repeated to obtain a prediction model for each road segment in each route.

[0009] The length of the period T is one week, and the length of each time period in the period T is equal, which is 1 hour.

[0010] S3. For each road segment in each route, the trained prediction model is used to predict traffic flow information and time information for passing through the road segment, and the route length, traffic flow information of each road segment in the route and time information for passing through the road segment are comprehensively considered for route selection to realize traffic management of vehicles.

[0011] The step S3 includes: S301. For any route, first, according to the vehicle departure time, the time period sequence number and date information in a period T where the departure time is located are determined; According to the date information of the vehicle departure time, the time period sequence number and the meteorological information of the departure time of the first road segment, the sample feature vector is constructed according to step S201, the prediction model of the first road segment in the current route is input, and the traffic flow information of the first road segment and the time information for passing through the first road segment are obtained; S303. On the basis of the vehicle departure time, the predicted time information of passing through the first section of road is added to obtain the departure time of the second section of road, thereby obtaining the time section information and date information of the second section of road, and the meteorological information corresponding to the departure time of the second section of road, combining the meteorological information, the date information and the time section serial number, the sample feature vector is constructed according to step S201, the prediction model of the second section of road in the current route is input, the traffic flow information of the second section of road and the time information of passing through the second section of road are obtained; Similarly, from the third section of road, the departure time of each section of road and the departure time of passing through the previous section of road are added to obtain the predicted time of passing through the previous section of road, and the date information and the meteorological information of the current section of road at the departure time are obtained through the obtained departure time; and the sample feature vector is constructed to predict the traffic flow information and the time information of passing through the road of each section of road in turn until the prediction of the last section of road is completed; S304. For each route, steps S301-S303 are repeatedly executed to complete the traffic flow information and the time information of passing through the road of each section of road in each route; S305. The route selection is performed by comprehensively considering the route length, the traffic flow information of each section of road in the route and the time of passing through the road section, and the traffic management of the vehicle is realized.

[0012] The meteorological information in step S3 is obtained through weather forecast.

[0013] The step S305 comprises: For each route, the average traffic flow is calculated according to the predicted traffic flow information of each section of road , the time information of passing through the road section is added to obtain the total time of passing through the route , the total mileage of the route , the priority coefficient of the route is obtained by weighting , and the route with the largest priority coefficient is selected as the traffic route of the vehicle.

[0014] The above description shows and describes one preferred embodiment of the present application, but as previously mentioned, it should be understood that the present application is not limited to the form disclosed herein, should not be considered as excluding other embodiments, and can be used in various other combinations, modifications and environments, and can be modified by the above teachings or related art or knowledge within the scope of the inventive concept described herein. Any modification and change made by those skilled in the art without departing from the spirit and scope of the present application shall be within the protection scope of the appended claims of the present application.

Claims

1. A road vehicle traffic management method based on data prediction, characterized in that: Includes the following steps: S1. Given a vehicle's starting point and destination, and multiple routes from the starting point to the destination, each route contains multiple road segments; S2. Collect historical traffic flow information, historical weather information, historical date information, and time information for each segment of each route, and build and train a prediction model for each segment based on the collected information; S3. For each road segment in each route, use the trained prediction model to predict traffic flow information and the time taken to pass through the road. Combine the route length, traffic flow information of each road segment in the route and the time taken to pass through the road segment to select the route and realize vehicle traffic management.

2. The road vehicle traffic management method based on data prediction according to claim 1, characterized in that: Step S2 includes: S201. For any road segment, collect historical traffic flow information, time information taken to pass through the road segment, historical weather information, and historical date information within period T, and construct training samples for the current period; If the period contains N time intervals, then: The historical traffic flow information is recorded as follows: ,in This represents the historical traffic flow information within the nth time period, where n is the time period number, n=1,2,…,N; The time taken to travel the road is recorded as follows: ,in This represents the time taken to travel through the road during the nth time period, where n = 1, 2, ..., N; The average time taken by a number of vehicles to pass through the road segment during the nth time period is obtained. The historical meteorological information is denoted as , n=1,2,…,N; where This represents historical meteorological information within the nth time period, where... This represents the average temperature over the nth time period. This represents the weather information for the nth time period. Indicates a cloudy day. Indicates sunny or cloudy weather. It means rain or snow. This indicates extreme weather conditions, including hail, tornadoes, and thunderstorms; The historical date information is denoted as , n=1,2,…,N; where This represents historical date information within the nth time period. Indicates working days. Indicates a non-working day; Construct N training samples for the current period, where the training sample for the nth time period is denoted as . , n=1,2,…,N, where, For sample feature vectors, The sample label vector; S202. Repeat step S201 within multiple periods T to obtain training samples within multiple periods T and add them to the training sample set of the current road segment; S203. Construct a prediction model for the current road segment based on a deep neural network algorithm, and train the prediction model using the training sample set of the current road segment to obtain a trained prediction model. S204. For each road segment in each route, repeat steps S201 to S203 to obtain the prediction model for each road segment in each route.

3. The road vehicle traffic management method based on data prediction according to claim 1, characterized in that: The duration of period T is one week, and the duration of each time segment within period T is equal, which is 1 hour.

4. The road vehicle traffic management method based on data prediction according to claim 1, characterized in that: Step S3 includes: S301. For any route, first determine the time period number and date information within a period T based on the vehicle's departure time; By combining the date information of the vehicle's departure time, the time period number, and the meteorological information of the departure time of the first road segment, a sample feature vector is constructed according to step S201. This vector is then input into the prediction model of the first road segment in the current route to obtain the traffic flow information of the first road segment and the time information taken to pass through the first road segment. S303. Based on the vehicle's departure time, add the predicted time information for passing through the first road segment to obtain the departure time of the second road segment, thereby obtaining the time period information and date information of the second road segment, as well as the meteorological information corresponding to the departure time of the second road segment. Combining the meteorological information, date information, and time period sequence number, construct the sample feature vector according to step S201, input it into the prediction model of the second road segment in the current route, and obtain the traffic flow information of the second road segment and the time information for passing through the second road segment. Similarly, starting from the third road segment, the departure time of each road segment and the departure time required to pass through the previous road segment are added to the predicted time required to pass through the previous road segment. The date information and the meteorological information of the current road segment at the departure time are obtained through the obtained departure time. A sample feature vector is constructed to predict the traffic flow information and the time required to pass through each road segment in turn, until the prediction of the last road segment is completed. S304. For each route, repeat steps S301 to S303 to complete the traffic flow information and the time taken to pass through each road segment in each route; S305. The route selection is based on a combination of route length, traffic flow information of each road segment, and time taken to pass through each segment, thereby enabling vehicle traffic management.

5. The road vehicle traffic management method based on data prediction according to claim 4, characterized in that: Step S305 includes: For each route, the average traffic flow is calculated based on the predicted traffic flow information for each road segment. The predicted travel time for each road segment is then summed to obtain the total travel time for the route. Then the total mileage of the route The priority coefficient of the route is obtained by weighting. Then, the route with the highest priority coefficient is selected as the vehicle's travel route.

Images