KR102384433B1 - Method of creating accuracy evaluation model for traffic information and method of providing traffic information using the same - Google Patents

Abstract

The present invention relates to a method for generating a traffic information accuracy evaluation model and a method for providing traffic information using a traffic information accuracy evaluation model, the method comprising: collecting, by a server, a speed distribution of a probe vehicle for each preset speed range; Separating the speed distribution for each speed range into at least two or more sections, and generating, by the server, an accuracy evaluation model according to the section divided for each speed range.

Description

A method of generating a traffic information accuracy evaluation model and a method of providing traffic information using a traffic information accuracy evaluation model

The present invention relates to a method of generating a traffic information accuracy evaluation model and a method of providing traffic information using a traffic information accuracy evaluation model, and more particularly, a method of generating a traffic information accuracy evaluation model for evaluating the accuracy of the provided traffic information, and evaluating the accuracy of traffic information It relates to a method of providing traffic information using a model.

Recently, by providing section traffic information to road users in real time, technologies have been provided to minimize operation and time costs, and to prevent traffic concentration by distributing vehicles as much as possible to increase the efficiency of a limited road.

Such traffic information is generally calculated by a traffic information providing device based on information provided by a plurality of probe vehicles. Traffic information is often calculated.

The map data includes road links, a list of nodes for each link, and coordinates of each node. Conventional traffic information providing devices use the map data and information provided by the probe vehicle to determine the average speed or specific speed for each link. Traffic information was calculated by calculating the average speed for each section.

For example, by displaying blue information on a road with smooth traffic, yellow information on a delayed (or slow-moving) road, and red information on a congested road, a service was provided so that the user could check the road condition.

However, when driving on the road while receiving these services, the actual road condition is communicated smoothly, but the map is displayed as a delay condition. There was a problem that it occurred.

Meanwhile, the background technology of the present invention is disclosed in Korean Patent Laid-Open No. 10-2007-0042445 (2007.04.23).

It is an object of the present invention to provide a method for generating a traffic information accuracy evaluation model and a method for providing traffic information using the traffic information accuracy evaluation model, which can increase the accuracy of traffic information by evaluating the accuracy of the provided traffic information.

A method for generating a traffic information accuracy evaluation model according to the present invention includes: collecting, by a server, a speed distribution of a probe vehicle for each preset speed range; dividing, by the server, a speed distribution for each speed range into at least two or more sections; and generating, by the server, an accuracy evaluation model according to a section divided for each speed range.

In the step of classifying the speed distribution for each speed range of the present invention, the server classifies the speed distribution according to the frequency of the speed.

In the step of classifying the speed distribution for each speed range of the present invention, the server provides a section corresponding to the first accuracy, a section corresponding to a second accuracy that is a value smaller than the first accuracy, and a second value smaller than the second accuracy It is characterized in that the velocity distribution is divided into sections corresponding to 3 accuracy.

In the step of generating the accuracy evaluation model of the present invention, the server is characterized in that it generates the accuracy evaluation model through a regression line calculated based on the boundary speed of each section.

A method of providing traffic information using a traffic information accuracy evaluation model according to the present invention includes: collecting, by a server, GPS trajectory information of a probe vehicle; generating, by the server, a service rate based on the collected trajectory information; providing, by the server, traffic information according to the generated service speed; collecting, by the server, ground truth information of the evaluation section; and calculating, by the server, the accuracy of the evaluation section based on the true value information of the evaluation section, the service speed corresponding to each true value, and a pre-generated accuracy evaluation model.

In the present invention, the accuracy evaluation model is divided into at least two or more regions according to accuracy, and is characterized in that it is in the form of a plane with the true value and service speed as axes.

In the step of calculating the accuracy of the evaluation section of the present invention, the server is characterized in that it calculates the accuracy through the calculation shown in Equation 1 below.

The traffic information providing method using the traffic information accuracy evaluation model according to the present invention is characterized in that it further comprises the step of changing, by the server, the service speed generation method of the corresponding evaluation section when the calculated accuracy is less than a set value.

The traffic information providing method using the traffic information accuracy evaluation model according to the present invention is characterized in that it further comprises, before the step of collecting the trajectory information, the step of generating the accuracy evaluation model by the server.

In the present invention, generating the accuracy evaluation model includes: collecting, by the server, a speed distribution of the probe vehicle for each preset speed range; dividing, by the server, a speed distribution for each speed range into at least two or more sections; and generating, by the server, an accuracy evaluation model according to a section divided for each speed range.

According to the present invention, the traffic information accuracy evaluation model generation method and the traffic information provision method using the traffic information accuracy evaluation model evaluate the quality of traffic information in consideration of tolerance by selecting the accuracy according to the section divided by frequency and generating the evaluation model. has the effect of making it possible.

1 is an exemplary diagram for explaining an apparatus for performing a method for generating a traffic information accuracy evaluation model and a method for providing traffic information using a traffic information accuracy evaluation model according to an embodiment of the present invention.
2 is a flowchart illustrating a method of generating a traffic information accuracy evaluation model according to an embodiment of the present invention.
3 is an exemplary diagram for explaining a speed distribution of a probe vehicle for each speed range in a method for generating a traffic information accuracy evaluation model according to an embodiment of the present invention.
4 is an exemplary diagram for explaining a process of dividing a speed distribution into sections in a method for generating a traffic information accuracy evaluation model according to an embodiment of the present invention.
5 is an exemplary diagram for explaining an evaluation model generated in a method for generating a traffic information accuracy evaluation model according to an embodiment of the present invention.
6 is a flowchart illustrating a method of providing traffic information using a traffic information accuracy evaluation model according to an embodiment of the present invention.

Hereinafter, an embodiment of a method for generating a traffic information accuracy evaluation model and a method for providing traffic information using a traffic information accuracy evaluation model according to the present invention will be described with reference to the accompanying drawings. In this process, the thickness of the lines or the size of the components shown in the drawings may be exaggerated for clarity and convenience of explanation. In addition, the terms to be described later are terms defined in consideration of functions in the present invention, which may vary according to the intention or custom of the user or operator. Therefore, definitions of these terms should be made based on the content throughout this specification.

1 is an exemplary diagram for explaining an apparatus for performing a method for generating a traffic information accuracy evaluation model and a method for providing traffic information using a traffic information accuracy evaluation model according to an embodiment of the present invention.

As shown in FIG. 1 , the server 100 includes an information collection unit 110 , an accuracy evaluation model generation unit 120 , a service speed generation unit 130 , a traffic information providing unit 140 , and a traffic information evaluation unit ( 150) may be included.

That is, the server 100 may be a traffic information service server that collects road condition information from a plurality of probe vehicles, generates traffic information, and provides the generated traffic information to each vehicle that is actually driving. .

The information collection unit 110 may collect GPS trajectory information, speed information, and the like from a probe vehicle, a vehicle receiving traffic information, and the like. For example, the information collection unit 110 may collect information directly from each vehicle or collect information from a traffic information collection device installed on a road.

The accuracy evaluation model generation unit 120 may generate an accuracy evaluation model capable of evaluating the accuracy of traffic information by analyzing and processing the data collected and accumulated through the information collection unit 110 .

The service speed generating unit 130 may generate a service speed (representative speed calculated for providing traffic information) of each link or a specific section according to real-time information collected through the information collecting unit 110 . For example, the service speed generator 130 may calculate the average speed of the probe vehicles in a specific section collected during the set period as a representative speed (service speed) of the section.

The traffic information providing unit 140 may generate traffic information according to the service speed generated by the service speed generating unit 130 and provide it to each vehicle. For example, the traffic information providing unit 140 may provide traffic information such as delay or congestion in a specific section, service speed, and the like.

The traffic information evaluation unit 150 may evaluate the traffic information provided to each vehicle. That is, the traffic information evaluation unit 150 may evaluate the traffic information provided to each vehicle by using the accuracy evaluation model generated by the accuracy evaluation model generation unit 120 .

2 is a flowchart for explaining a method for generating a traffic information accuracy evaluation model according to an embodiment of the present invention, and FIG. 3 is a probe vehicle for each speed range in the method for generating a traffic information accuracy evaluation model according to an embodiment of the present invention. It is an exemplary view for explaining the speed distribution of , and FIG. 4 is an exemplary view for explaining the process of dividing the speed distribution into sections in the method of generating a traffic information accuracy evaluation model according to an embodiment of the present invention, and FIG. As an exemplary diagram for explaining the evaluation model generated in the method for generating a traffic information accuracy evaluation model according to an embodiment of the present invention, a method for generating a traffic information accuracy evaluation model according to the present embodiment will be described with reference to this.

As shown in FIG. 2 , the server 100 first collects the speed distribution of the probe vehicle for each speed range ( S200 ). That is, as can be seen in FIG. 3, even on a link (road) corresponding to a specific speed range (eg, 10 to 20 km/h), the actual speed of the probe vehicle passing through the link may have various values, and the server ( 100) may collect the speed distribution of the probe vehicle for each speed range. At this time, such a speed range may be preset, for example, may be set at an interval of 10 km/h as shown in FIG. 3 .

Next, the server 100 selects an accuracy section for each speed range based on the speed distribution collected in the step S200 (S210). That is, as can be seen in FIG. 4, the server 100 calculates the speed distribution for the range of 10 to 20 km/h in a plurality of sections (eg, 100% accuracy section, 50% accuracy section, and error section (0% accuracy section)) ) can be distinguished.

In this case, the server 100 may classify the speed distribution according to the frequency of the speed. For example, the server 100 selects a section including a frequency of 50% based on the speed corresponding to the mode as a 100% accuracy section, and selects a section including a frequency of 70% as a 50% accuracy section, , the remaining section can be selected as an error section.

The server 100 may select the section of the speed distribution in other speed ranges as well, so that the speed distribution for each speed range can be divided into a plurality of sections.

That is, in the prior art, it is common to evaluate the quality of the provided service speed using only an error ratio between the service speed provided to each vehicle and a ground truth (actual speed). However, in the case of this method, there were many problems in using it as analysis data because it was not possible to reflect the variation of errors that may occur depending on the characteristics of the road (link, evaluation section), speed range, etc.

Unlike the conventional method, in the present embodiment, the accuracy is selected by dividing the section of the speed distribution according to the frequency of the collected speed, so that a model capable of evaluating the accuracy by reflecting a realistic tolerance can be generated.

After the step (S220), the server 100 generates an accuracy evaluation model according to the accuracy section for each speed range selected in the step (S210) (S220). That is, the server 100 may generate an accuracy evaluation model according to a section divided for each speed range.

In this case, the server 100 may generate an accuracy evaluation model through a regression line calculated based on the boundary speed of each section. For example, as can be seen in FIGS. 4 and 5, in the range of 10 to 20 km/h, there is an upper (high speed) boundary speed and a lower (low speed) boundary speed that separates the 100% accuracy section and the 50% accuracy section. Similarly, in other speed ranges, there is an upper boundary speed and a lower boundary speed that divide the 100% accuracy section and the 50% accuracy section, respectively. That is, the boundary speed here means the speed at which each section is divided.

The server 100 calculates a regression line for the upper boundary velocities that separate the 100% accuracy section and the 50% accuracy section, and can calculate the upper boundary line that separates the 100% accuracy section and the 50% accuracy section, and In the same way, the lower boundary line that separates the 100% accuracy section and the 50% accuracy section can be calculated. In this case, the server 100 may calculate the regression line using the method of least squares.

영역에 속하게 되며, 이러한 경우 해당 서비스 속도는 50%의 정확도를 가지는 경우라고 볼 수 있다.The server 100 may generate an accuracy evaluation model in such a way that the entire area is divided into a plurality of areas through the calculated boundary lines. That is, when evaluating the accuracy using the accuracy evaluation model, the x-axis of FIG. 5 may be an axis corresponding to the true value, and the y-axis may be an axis corresponding to the service speed. In other words, if the service speed was 50 km, but the actual value was 30 km,

area, and in this case, the corresponding service speed can be regarded as having an accuracy of 50%.

영역의 경우 100% 정확도를 가지는 영역이라고 볼 수 있고,

영역의 경우 50%의 정확도를 가지는 영역이라고 볼 수 있으며,

영역의 경우 0%의 정확도를 가지는 영역이라고 볼 수 있다.That is, in Figure 5,

In the case of an area, it can be regarded as an area with 100% accuracy,

In the case of an area, it can be considered as an area with 50% accuracy,

In the case of an area, it can be regarded as an area having an accuracy of 0%.

On the other hand, the above-described accuracy evaluation model is preferably generated for each link (evaluation section) of the road, but may be generated only for the main link.

6 is a flowchart for explaining a method of providing traffic information using a traffic information accuracy evaluation model according to an embodiment of the present invention. Referring to this, the traffic information accuracy evaluation model according to this embodiment will be described as follows. In addition, the traffic information providing method to be described below is a traffic information providing method performed for one evaluation section (eg, a specific link), and traffic information for a plurality of roads may be provided at the same time.

6, the server 100 first generates an accuracy evaluation model (S300). For example, the server 100 collects the speed distribution of the probe vehicle for each preset speed range, divides the speed distribution for each speed range into at least two or more sections, and the accuracy according to the section divided for each speed range An evaluation model can be created.

Next, the server 100 collects GPS trajectory information of the probe vehicle (S310). For example, the server 100 may collect GPS trajectory information from a plurality of probe vehicles during a set period (information provision space interval).

After the step (S310), the server 100 generates a service rate based on the trajectory information collected in the step (S310) and provides it (S320). For example, the server 100 may generate a service speed by matching a GPS trajectory and a map, and calculating an average speed of probe vehicles traveling on a specific road (evaluation section).

After the step (S320), the server 100 collects ground truth information of the evaluation section (S330). Here, the true value means the actual speed of the vehicle that has driven the corresponding evaluation section.

Then, the server 100 calculates the accuracy of the evaluation section based on the true value information of the evaluation section collected in step S330, the service speed corresponding to each true value, and the accuracy evaluation model generated in the step S300. do (S340).

Specifically, the server 100 may calculate the accuracy of the corresponding evaluation section through the calculation shown in Equation 1 below.

는 수집된 참값 중 정확도 평가 모형의 제1영역에 해당하는 데이터의 수,

는 수집된 참값 중 정확도 평가 모형의 제2영역에 해당하는 데이터의 수,

은 수집된 참값 중 정확도 평가 모형의 제n영역에 해당하는 데이터의 수,

는 수집된 전체 참값의 수를 의미한다.where Acc is the accuracy (%), a ₁ , a ₂ and a _n are constants according to the accuracy of each area of the accuracy evaluation model,

is the number of data corresponding to the first area of the accuracy evaluation model among the collected true values;

is the number of data corresponding to the second area of the accuracy evaluation model among the collected true values;

is the number of data corresponding to the n-th area of the accuracy evaluation model among the collected true values;

is the total number of true values collected.

For example, when using the accuracy evaluation model as shown in FIG. 5 , the server 100 may calculate the accuracy of the evaluation section through the calculation shown in Equation 2 below.

영역의 경우 100% 정확도를 가지는 영역이므로 상수로 1을 곱하면 되고,

영역의 경우 50%의 정확도를 가지는 영역이므로 상수로 0.5를 곱하면 되며,

영역의 경우 0%의 정확도를 가지므로 해당 항목은 소거되게 된다.In other words

In the case of an area, it is an area with 100% accuracy, so you can multiply it by 1 by a constant,

In the case of an area, it is an area with 50% accuracy, so multiply by 0.5 by a constant.

In the case of an area, the item is deleted because it has 0% accuracy.

영역에 속하게 되므로,

에 해당하는 값이 되며, 서버(100)는 이러한 방식으로 해당 평가구간의 정확도를 산출할 수 있다.That is, if the service speed was 50 km, but the actual value was 30 km,

Because it belongs to the realm

is a value corresponding to , and the server 100 may calculate the accuracy of the corresponding evaluation section in this way.

영역이나

영역에 많이 포함되어 있는 경우에는, 서버(100)가 제공한 서비스 속도가 과도한 값으로 제공되었다고 볼 수 있으므로, 서버(100)는 해당 평가구간의 서비스 속도 생성 시 보다 작은 값이 생성되도록 그 생성 방식을 변경할 수 있다.On the other hand, when the calculated accuracy is less than the set value, the server 100 may change the method of generating the service speed of the corresponding evaluation section by the server. For example, if the distribution of true values is

area or

When a large number is included in the area, since it can be considered that the service speed provided by the server 100 is provided as an excessive value, the server 100 generates a smaller value when generating the service speed of the corresponding evaluation section. can be changed

As such, the method of generating the traffic information accuracy evaluation model and the method of providing traffic information using the traffic information accuracy evaluation model according to the embodiment of the present invention select the accuracy according to the section divided by frequency and generate the evaluation model, taking into account the tolerance. It is possible to evaluate the quality of information and to analyze the cause of the decrease in the accuracy of traffic information, thereby increasing the accuracy of traffic information.

Although the present invention has been described with reference to the embodiment shown in the drawings, this is merely exemplary, and it is understood that various modifications and equivalent other embodiments are possible by those of ordinary skill in the art. will understand Therefore, the technical protection scope of the present invention should be defined by the following claims.

100: server
110: information collection unit
120: Accuracy evaluation model generation unit
130: service speed generation information
140: traffic information provider
150: Traffic Information Evaluation Department

Claims (13)

delete delete collecting, by a server, GPS trajectory information of a probe vehicle;
generating, by the server, a service rate based on the collected trajectory information;
providing, by the server, traffic information according to the generated service speed;
collecting, by the server, ground truth information of the evaluation section; and
Comprising the step of the server calculating the accuracy of the evaluation section based on the true value information of the evaluation section, the service speed corresponding to each true value, and a pre-generated accuracy evaluation model,
Before the step of collecting the trajectory information, the server further comprises the step of generating the accuracy evaluation model,
The step of generating the accuracy evaluation model comprises:
collecting, by the server, a speed distribution of the probe vehicle for each preset speed range;
dividing, by the server, a speed distribution for each speed range into at least two or more sections; and
Including, by the server, generating an accuracy evaluation model according to a section divided for each speed range,
In the step of classifying the speed distribution for each speed range, the server determines the section corresponding to the first accuracy, the section corresponding to the second accuracy that is a value smaller than the first accuracy, and the third accuracy that is a value smaller than the second accuracy. A method of providing traffic information using a traffic information accuracy evaluation model, characterized in that the speed distribution is divided into corresponding sections.
collecting, by a server, GPS trajectory information of a probe vehicle;
generating, by the server, a service rate based on the collected trajectory information;
providing, by the server, traffic information according to the generated service speed;
collecting, by the server, ground truth information of the evaluation section; and
Comprising the step of the server calculating the accuracy of the evaluation section based on the true value information of the evaluation section, the service speed corresponding to each true value, and a pre-generated accuracy evaluation model,
Before the step of collecting the trajectory information, the server further comprises the step of generating the accuracy evaluation model,
The step of generating the accuracy evaluation model comprises:
collecting, by the server, a speed distribution of the probe vehicle for each preset speed range;
dividing, by the server, a speed distribution for each speed range into at least two or more sections; and
Including, by the server, generating an accuracy evaluation model according to a section divided for each speed range,
In the step of generating the accuracy evaluation model, the server generates the accuracy evaluation model through a regression line calculated based on the boundary speed of each section.
collecting, by a server, GPS trajectory information of a probe vehicle;
generating, by the server, a service rate based on the collected trajectory information;
providing, by the server, traffic information according to the generated service speed;
collecting, by the server, ground truth information of the evaluation section; and
Comprising the step of the server calculating the accuracy of the evaluation section based on the true value information of the evaluation section, the service speed corresponding to each true value, and a pre-generated accuracy evaluation model,
The accuracy evaluation model is divided into at least two or more regions according to accuracy, and the traffic information providing method using the traffic information accuracy evaluation model, characterized in that it is in the form of a plane with the true value and service speed as axes.
delete 6. The method of claim 5,
In the step of calculating the accuracy of the evaluation section, the server is a traffic information providing method using a traffic information accuracy evaluation model, characterized in that for calculating the accuracy through the calculation shown in Equation 1 below.
(Equation 1)

(Where Acc is the accuracy (%), a ₁ , a ₂ and a _n are constants according to the accuracy of each area of the accuracy evaluation model,

is the number of data corresponding to the first area of the accuracy evaluation model among the collected true values;

is the number of data corresponding to the second area of the accuracy evaluation model among the collected true values;

is the number of data corresponding to the n-th area of the accuracy evaluation model among the collected true values;

is the total number of true values collected.)
6. The method of claim 5,
When the calculated accuracy is less than the set value, the method of providing traffic information using the traffic information accuracy evaluation model, characterized in that the server further comprises the step of changing the service speed generation method of the evaluation section.
6. The method of claim 5,
Before the step of collecting the trajectory information, the method of providing traffic information using the traffic information accuracy evaluation model, characterized in that it further comprises the step of generating the accuracy evaluation model by the server.
10. The method of claim 9,
The step of generating the accuracy evaluation model comprises:
collecting, by the server, a speed distribution of the probe vehicle for each preset speed range;
dividing, by the server, a speed distribution for each speed range into at least two or more sections; and
and generating, by the server, an accuracy evaluation model according to a section divided for each speed range.
11. The method of claim 10,
In the step of classifying the speed distribution for each speed range, the server classifies the speed distribution according to the frequency frequency of the speed.
11. The method of claim 10,
In the step of classifying the speed distribution for each speed range, the server determines the section corresponding to the first accuracy, the section corresponding to the second accuracy that is a value smaller than the first accuracy, and the third accuracy that is a value smaller than the second accuracy. A method of providing traffic information using a traffic information accuracy evaluation model, characterized in that the speed distribution is divided into corresponding sections.
11. The method of claim 10,
In the step of generating the accuracy evaluation model, the server generates the accuracy evaluation model through a regression line calculated based on the boundary speed of each section.

Images