KR20210096464A - Appartus for controlling traffic information, method thereof and storage media storing a program for controlling traffic information - Google Patents

Abstract

The present invention relates to a traffic information control apparatus, a traffic information control method, and a storage medium storing a program for controlling traffic information, which can intuitively recognize the communication situation of roads and control a traffic situation. According to an embodiment of the present invention, the traffic information control method comprises: a step of receiving traffic data on roads connected to a crossroad; a step of preprocessing the received traffic data values on the roads; a step of generating a polygon image including the crossroad and having vertices corresponding to the preprocessed traffic data values; a step of performing a deep learning algorithm with the generated polygon image as input data; and a step of providing traffic signal control information in accordance with a result of performing the deep learning algorithm.

Description

A storage medium for storing a traffic information control device, a traffic information control method, and a program for controlling traffic information

The following disclosure relates to a traffic information control device, a traffic information control method, and a storage medium for storing a program for controlling traffic information.

Traffic data had difficulties in predicting and expressing it due to characteristics such as complexity according to traffic flow and congestion according to traffic conditions.

Artificial intelligence models have shown excellent results for predicting data showing complex patterns.

However, in the case of the existing traffic information, there was a limit to using such an artificial intelligence model because the traffic situation was expressed by the expression of color on the road. In particular, unlike highways, the urban road network is connected in a mesh shape, so it was difficult to intuitively recognize the urban communication situation.

In the intelligent transportation system that expresses the traffic situation, the traditional method was a link-based expression method.

According to this method, the speed of travel on the road on the map is displayed in color, and the speed of travel on the road in the opposite direction is also displayed in color in many cases. The structure and characteristics of these data had a problem that it was difficult to be used as an artificial intelligence model.

In the case of urban traffic conditions, the traffic conditions of vehicles are expressed based on a link with clear starting and ending points. Unlike highways, vehicles can start or stop at any point in this way of expression, but they are not reflected in the link-based travel speed, resulting in errors in traffic prediction.

The existing method of expressing traffic conditions is to display the speed of traffic on the road on the map in color. Since the traffic situation of the two-way road is displayed on the map, this method is difficult for artificial intelligence to learn.

In addition, since the existing method provides information on the traffic situation centered on the main road, there is a problem in that traffic congestion spreads and it is difficult to express the interaction between the roads.

An object of the following disclosure is to provide a traffic information control device capable of intuitively recognizing the traffic situation of a road and controlling the traffic situation, a traffic information control method, and a storage medium for storing a program for controlling the traffic information .

Another object of the present disclosure is to provide a traffic information control device, a traffic information control method, and a storage medium for storing a program for controlling traffic information so that an artificial intelligence model can be optimally used to control traffic information.

Another object of the present disclosure is to provide a traffic information control device capable of controlling traffic information by intuitively expressing traffic conditions or interaction of roads, a traffic information control method, and a storage medium for storing a program for controlling traffic information will do

An embodiment includes the steps of: receiving traffic data on roads connected to an intersection; preprocessing the received traffic data values on the roads; generating a polygonal image including the intersection and having vertices corresponding to the values of the preprocessed traffic data; performing a deep learning algorithm using the generated polygonal image as input data; and providing traffic signal control information according to the execution result of the deep learning algorithm.

The deep learning algorithm may be a Convolutional Neural Network (CNN) algorithm.

The traffic data value may include vehicle speed information on the roads.

The polygonal image may have the intersection as a center and directions of the roads corresponding to the positions of the vertices.

Distances from the center to the vertices of the polygonal image may each be inversely proportional to vehicle speeds on the roads.

A distance from the center to a first vertex among the vertices may be proportional to the degree of traffic congestion between the center and the first road corresponding to the vertex.

The polygonal image may have a maximum distance from the center of the polygonal image to a first vertex corresponding to the first road when the degree of traffic congestion on a first road among the roads is maximum.

When the degree of traffic congestion on the second road is the minimum, the distance from the center of the polygonal image to the first vertex corresponding to the second road may be the minimum.

Another embodiment provides a receiving module for receiving traffic data on roads connected to an intersection; a preprocessing module for preprocessing the received traffic data values on the roads; a data generating module for generating a polygonal image including the intersection and having vertices corresponding to the values of the preprocessed traffic data; an algorithm processing module for performing a deep learning algorithm based on the generated polygonal image; and a control module that provides traffic signal control information according to a result of performing the deep learning algorithm.

Another embodiment, the receiving unit for receiving the traffic data on the roads connected to the intersection of the roads; Preprocessing the received traffic data values on the roads, generating a polygonal image including the intersection and having vertices corresponding to the preprocessed traffic data values, and deep learning based on the generated polygonal image a processor for performing (deep learning) algorithms; and a control unit configured to provide traffic signal control information according to a result of performing the deep learning algorithm.

Another embodiment receives traffic data on roads connected to an intersection; Preprocessing the received traffic data values on the roads, generating a polygonal image including the intersection and having vertices corresponding to the preprocessed traffic data values, and deep learning based on the generated polygonal image (deep learning) performing an algorithm, and accurately determining the traffic conditions of the roads according to the execution result of the deep learning algorithm; A storage medium for storing a program for controlling traffic information is provided.

According to the following embodiment, it is possible to intuitively recognize the traffic situation of the road and control the traffic situation.

According to the following embodiments, an artificial intelligence model can be optimally used to control traffic information, so that traffic information can be efficiently controlled.

According to the following embodiment, traffic information can be easily controlled by intuitively expressing traffic conditions or interaction of roads.

1 is a view showing an example of a traffic information control method according to an embodiment;
2 is a diagram illustrating an example of generating a polygon image from traffic data of an intersection
3 is a view showing an embodiment of a traffic signal control device;
4 is a view showing another embodiment of a traffic signal control device;
5 is a view showing an example of an image of a traffic jam situation according to the embodiment;

Hereinafter, embodiments will be described in detail with reference to the drawings.

1 is a diagram illustrating an example of a traffic information control method according to an embodiment.

Receives traffic data on roads connected to the intersection (S100).

Receive traffic data on the road in real-time or near real-time. It is possible to receive the speed of vehicles on the road and the like based on the intersection at the point where the roads meet. It is possible to determine a congestion situation on the road from traffic data including vehicle speed.

The received traffic data values of roads are pre-processed (S200).

The traffic data received as described above may be pre-processed by matching units, etc. so that it can be processed in the embodiment, and determining representative data when there are various types of data on the road.

A polygonal image including the intersection and having vertices corresponding to the values of the preprocessed traffic data is generated (S300).

In the polygonal image including the intersection on the road, the distance between the vertex and the center of the polygon may be changed according to the speed of the vehicle.

For example, the center of the polygon may be an intersection, and respective distances from the intersection to the vertex may be inversely proportional to the speed of a vehicle on roads connected to the intersection.

In other words, the distances from the center to the vertices of the polygonal image may be inversely proportional to the vehicle speeds on the roads, respectively.

That is, the intersection may be the center of the polygon, and the length of the first vertex from the center of the polygon may correspond to the vehicle speed on the first road connected to the intersection.

When the vehicle speed on the first road is low (when the vehicle is close to congestion), the length of the first vertex from the center of the polygon may be increased.

When the vehicle speed on the first road is high (when the vehicle speed is close to smooth), the length of the first vertex from the center of the polygon may be shortened.

Therefore, when the polygon is observed from above, the point where the vehicle speed is high for the polygon on the intersection is short from the center to the vertex of the polygon, and when the vehicle speed is slow, the distance from the center to the vertex of the polygon is long.

As described above, the line from the center of the polygon to the vertex may correspond to the direction of the road.

The distance from the center of the polygon to the vertex may correspond to a value of traffic data, such as a speed of a vehicle on the road.

As another example, the distance from the center of the polygon to the first vertex among the vertices may be proportional to the degree of traffic congestion of the first road corresponding to the center and the vertex.

An example for a polygonal image is exemplified in detail below.

The polygons generated at each intersection according to the above method are input to the deep learning algorithm (S400).

The deep learning algorithm identifies polygonal images that are close to the input polygonal image.

For example, a deep learning algorithm may use a Convolutional Neural Network (CNN) algorithm.

According to the result of the deep learning algorithm, it is possible to accurately identify the traffic jam situation corresponding to the identified polygonal image. Traffic signal control information according to each traffic jam situation may be pre-stored.

Therefore, the result data according to the deep learning algorithm based on polygonal images can accurately determine various traffic jam situations.

Traffic signal control information is provided according to traffic congestion conditions determined according to the execution result of the deep learning algorithm (S500).

Current traffic signal control can be performed by outputting traffic signal control information corresponding to a traffic jam situation according to the result of the deep learning algorithm.

In particular, the traffic lights of several intersections that are interlocked are not easy to operate in conjunction unless the overall congestion situation is accurately determined.

However, as in the embodiment, if an accurate congestion situation is determined according to the execution result of the deep learning algorithm and traffic signal control information is provided, it is possible to efficiently control congestion linked to multiple intersections.

2 illustrates an example of generating a polygon image from traffic data of an intersection. An example of generating a polygon image from traffic data of an intersection will be described with reference to this drawing.

In this figure, (a) shows an example of the principle of generating a polygon image by using the degree of congestion among the traffic data on the intersection 10 of the road and the roads passing the intersection.

Line segments of east, west, north and south (represented by E, W, S, and N, respectively) passing through the intersection 10 correspond to roads leading to the intersection 10 , respectively. This example exemplifies an intersection in which directions of roads passing through the intersection 10 consist of east, west, north, and south for convenience of description.

The polygon image allows vertices to be created in the direction of the roads. Therefore, this polygon image has vertices in the east, west, south, and north directions.

The polygon image may have a maximum size when the degree of congestion of each road reaches a maximum (maximum traffic congestion), and may have a minimum size when the degree of congestion on each road is all minimum (minimum traffic congestion).

In other words, the polygonal image may be constructed such that the distances from the center of the polygonal image to the vertices are inversely proportional to vehicle speeds on the roads, respectively.

In this figure, (b) represents that the position of the vertex of the polygonal image changes according to the degree of congestion of the road in each direction from the intersection.

For example, when the traffic congestion of the intersection and the road in the east direction is maximum, the vertex corresponding to the east direction is located farthest from the center of the polygon image.

For example, if the traffic jam at the intersection and the road in the south direction is congested, the vertex corresponding to the south direction is located between the maximum distance and the minimum distance from the center of the polygon image according to the degree of congestion.

The position of the vertex may be positioned in proportion to the degree of congestion on each road.

That is, the distance from the center of the polygon to the first vertex among the vertices may be proportional to the degree of traffic congestion of the first road corresponding to the center and the first vertex.

In the polygonal image, when the degree of traffic congestion on the first road among the roads is the maximum, the distance from the center of the polygonal image to the first vertex corresponding to the first road may be maximized.

In the polygonal image, when the degree of traffic congestion on the second road is the minimum, the distance from the center of the polygonal image to the second vertex corresponding to the second road may be minimized.

The construction method of the polygon image in the drawing is an example.

Polygon images can be created differently according to the intention of the operator. For example, when the traffic congestion on the road in each direction is maximum, a polygon image having a minimum size may be generated, and when the traffic congestion on the road in each direction is minimum, a polygon image having the maximum size may be generated.

In addition, the center of the polygon image does not necessarily have to be an intersection point, and if the polygon image includes the intersection according to the shape and configuration of the intersection, an embodiment of the same concept can be configured sufficiently.

3 discloses an embodiment of a traffic signal control device. An embodiment of the traffic signal control device will be described with reference to this drawing.

One embodiment of the apparatus for controlling traffic signals includes a receiving module 110 , a preprocessing module 120 , a data generating module 130 , an algorithm processing module 140 , and a control module 150 .

The receiving module 110 receives traffic data on roads connected to the intersection of roads.

The pre-processing module 120 pre-processes the traffic data including the speed of vehicles on the roads received by the receiving module 110 .

The data generation module 130 generates a polygon image using pre-processed traffic data centered on the intersection. For example, the polygon image may be generated to include vertices and intersections corresponding to preprocessed traffic data values.

An example of generating a polygon image has been described above.

The algorithm processing module 140 performs a deep learning algorithm using the generated polygonal image as input data. The deep learning algorithm may use a Convolutional Neural Network (CNN) algorithm, etc., and may modify the CNN algorithm or use any algorithm used in the deep learning algorithm.

The algorithm processing module 140 may receive a polygonal image and accurately determine a current traffic jam situation according to the received polygonal image.

The algorithm processing module 140 may accurately determine not only the traffic congestion situation for a specific intersection, but also the congestion situation of several intersections that are interlocked and controlled.

The control module 150 may provide traffic signal control information corresponding to the traffic jam situation determined according to the execution result of the deep learning algorithm.

4 discloses another embodiment of a traffic signal control device. Another embodiment of the traffic signal control device will be described with reference to this drawing.

This drawing is a diagram showing the physical configuration of an embodiment of the traffic signal control device more clearly.

An embodiment of the apparatus for controlling traffic signals may include a receiver 210 , a processor 220 , and a control signal output unit 230 .

The receiver 210 receives traffic data on the intersection of roads or roads connected to the intersection. The traffic data may include vehicle speed and the like, and data capable of representing a real-time congestion situation may be included in the received data.

The processor 220 pre-processes the traffic data received by the receiver 210 and generates a polygon image corresponding to the intersection as described above.

The polygon image generated based on the traffic data may include an intersection and have vertices in each road direction.

As illustrated in this figure, a polygon image is made to correspond to each intersection, and the processor 220 generates polygon images corresponding to several intersections.

The processor 220 may receive polygon images, perform a deep learning algorithm, and accurately determine congestion conditions of corresponding intersections based on the input polygon images.

The processor 220 determines and outputs a congestion situation corresponding to a plurality of intersections according to a result of performing the deep learning algorithm. The processor 220 may determine the congestion situation of a plurality of intersections, intuitively recognize the traffic situation of the road, and control the congestion situation with an optimal traffic signal.

The control signal output unit 230 outputs a control signal corresponding to the congestion situation of the intersections received from the processor 220 .

Accordingly, the embodiment can intuitively express the traffic situation or the interaction of the road and efficiently control the traffic signal using the optimal traffic information.

5 shows an example in which a traffic jam situation is imaged according to an embodiment.

This figure shows the results (b) of performing a deep learning algorithm to generate a polygon image according to the actual traffic conditions of specific roads in Daejeon (a), and to determine an accurate congestion situation accordingly.

In this drawing, the left side shows the traffic situation at dawn (off-peak) when traffic is idle, and the right side shows the congestion situation during rush hour in the morning.

This figure shows an example of generating a polygonal image corresponding to an intersection on a real road connected to the south (S) and north (N) (a) and an example showing the result of performing a deep learning algorithm according to the generated polygonal image (b) represent each.

On the north-south roads, the circles indicate the intersections, and the congestion situation at the intersection is shaded in the polygon image. The polygon images corresponding to the intersection were expressed by overlapping the top and bottom along the north-south direction of the road.

For example, if a deep learning algorithm is performed on a polygon image (a) based on real-time received traffic data during non-peak or morning peak hours, congestion situations as shown in (b) can be determined.

For the images (b) in which the congestion situation is determined for each time period, a control signal capable of resolving each congestion situation may be output according to a traffic control algorithm already prepared.

Therefore, according to the embodiment, it is possible to intuitively recognize the traffic situation on the road and control the traffic situation. And, according to the embodiment, it is possible to easily determine the interaction of the road and to control it efficiently.

And according to the embodiment, it is possible to accurately determine the congestion situation for several intersections using the artificial intelligence model, and accordingly, an optimal control signal can be output, so that the congestion situation can be quickly controlled.

According to the embodiment, even when traffic congestion spreads and it is difficult to express the interaction between roads, it is possible to accurately determine and control the traffic situation.

110: receiving module
120: pre-processing module
130: data generation module
140: algorithm processing module
150: control module
210: receiver
220: processor
230: control signal output unit

Claims (18)

receiving traffic data on roads connected to the intersection;
preprocessing the received traffic data values on the roads;
generating a polygonal image including the intersection and having vertices corresponding to the values of the preprocessed traffic data;
performing a deep learning algorithm using the generated polygonal image as input data; and
Traffic information control method comprising a; providing traffic signal control information according to the execution result of the deep learning algorithm. The method of claim 1,
The deep learning algorithm is a traffic information control method that is a Convolutional Neural Network (CNN) algorithm. The method of claim 1,
The traffic data value includes vehicle speed information on the roads. The method of claim 1,
The polygonal image is
A traffic information control method that centers on the intersection and makes the directions of the roads correspond to the positions of the vertices. The method of claim 1,
and the distances from the center to the vertices of the polygonal image are inversely proportional to vehicle speeds on the roads, respectively. The method of claim 1,
The distance from the center to a first vertex among the vertices is proportional to the degree of traffic congestion of the first road corresponding to the center and the vertex. The method of claim 1,
The polygonal image is
When the degree of traffic congestion on the first road among the roads is maximum,
The distance from the center of the polygonal image to the first vertex corresponding to the first road is maximum. The method of claim 1,
The polygonal image is
When the degree of traffic congestion on the second road is minimal,
A method for controlling traffic information in which a distance from the center of the polygonal image to a first vertex corresponding to the second road is minimized. a receiving module for receiving traffic data on roads connected to the intersection;
a preprocessing module for preprocessing the received traffic data values on the roads;
a data generating module for generating a polygonal image including the intersection and having vertices corresponding to the values of the preprocessed traffic data;
an algorithm processing module for performing a deep learning algorithm based on the generated polygonal image; and
Traffic information control device comprising a; a control module for providing traffic signal control information according to the execution result of the deep learning algorithm. 10. The method of claim 9,
The polygonal image is
A traffic information control device centered on the intersection and making the directions of the roads correspond to the positions of the vertices. 10. The method of claim 9,
and the distances from the center to the vertices of the polygonal image are inversely proportional to vehicle speeds on the roads, respectively. 10. The method of claim 9,
A distance from the center to a first vertex among the vertices is proportional to the degree of traffic congestion of the first road corresponding to the center and the first vertex. 10. The method of claim 9,
The polygonal image is
When the degree of traffic congestion on the first road among the roads is maximum,
The distance from the center of the polygonal image to the first vertex corresponding to the first road is the maximum. 10. The method of claim 9,
The polygonal image is
When the degree of traffic congestion on the second road is minimal,
The distance from the center of the polygonal image to the first vertex corresponding to the second road is minimized. 10. The method of claim 9
The deep learning algorithm is a traffic information control device that is a Convolutional Neural Network (CNN) algorithm. 10. The method of claim 9,
The traffic data value includes vehicle speed information on the roads. a receiving unit for receiving traffic data on the roads connected to the intersection of the roads;
Preprocessing the received traffic data values on the roads, generating a polygonal image including the intersection and having vertices corresponding to the preprocessed traffic data values, and deep learning based on the generated polygonal image a processor for performing (deep learning) algorithms; and
Traffic information control device comprising a; a control unit for providing traffic signal control information according to the execution result of the deep learning algorithm. receive traffic data on roads connected to the intersection; Preprocessing the received traffic data values on the roads, generating a polygonal image including the intersection and having vertices corresponding to the preprocessed traffic data values, and deep learning based on the generated polygonal image (deep learning) performing an algorithm, and accurately determining the traffic conditions of the roads according to the execution result of the deep learning algorithm; A storage medium storing a program for controlling traffic information.

Images