KR20160036710A - System for predicting future traffic situations - Google Patents

Abstract

According to the present invention, it is possible to receive and store current traffic situation information in various road sections including a predetermined road section in real time, and to store construction information, event information and weather information Future event information and unexpected situation information from an external traffic control center server configured to receive and store future event information and unexpected situation information including the current traffic situation information, A future traffic situation prediction server is provided for predicting future traffic situation in the future. The server includes a database server configured to receive, process, and store current traffic situation information, future event information, and unexpected situation information for each road section from the external traffic control center server, And a traffic condition prediction module for predicting a traffic situation when a future event occurs in a predetermined road section and providing a predicted result to the server operator, System, the data of the database server is referred to based on various traffic condition-related parameters inputted through the user interface module, and a short-term traffic situation from present to one hour and a long-term traffic Analyzing the Situation and Example It comprises a traffic prediction system configured to be characterized.

Description

{SYSTEM FOR PREDICTING FUTURE TRAFFIC SITUATIONS}

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a traffic situation prediction system, and more particularly, to a traffic situation prediction system capable of predicting a future traffic situation using traffic information collected at an ITS (Intelligent Transportation System) center of a local government .

It is impossible to construct and expand road infrastructure in metropolitan areas where huge social and economic costs are generated due to traffic congestion in terms of budget constraint and road site securing. Therefore, pre-emptive traffic control and traffic demand dispersion strategy based on traffic situation prediction information are required to utilize the facilities already built as efficiently as possible.

In this regard, there is a system for predicting and indicating to what degree each road will become stagnant during the Chuseok holiday season in connection with a specific event such as Chuseok (for example, Japanese Patent Application Laid-Open No. 10-2004-86422). However, this prediction system predicts traffic conditions during the future Chuseok holiday season based on past data such as the speed and traffic volume collected in the past, for example, the Chuseok holiday season, and predicts future events (for example, Weather changes, road construction in characteristic road sections, events, unexpected situations, etc.).

These problems are also encountered in a real-time traffic situation information providing system in which demand is high. For example, on a highway such as a riverside north road, time information (speed information) required to pass through each section is provided, for example, by detecting the vehicle speed through a detector installed in each section. For example, when a traffic accident occurs in the vicinity of the Wonhyo Bridge in the above section, if the time passing through the Wonhyo Bridge from the Seongsu Bridge is 10 minutes in general, it is estimated that it takes about 30 minutes, for example, and it is provided to the driver through the sign. However, when the above-mentioned traffic accident occurs, it is predicted from the experience of the operator that the persons concerned or the like will be sent to the site or the CCTV, Can be collected at the headquarters of the traffic situation system. However, according to the present system, the operator of the system predicts that the time from the Seongsu Bridge to the Wonhyo Bridge will increase to about 30 minutes in light of the past experience, and provides the driver with the time to the driver, It is not provided to the driver with accurate time information as to whether or not it takes a substantial amount of time to pass.

Further, it is assumed that a user in Gangnam, for example, searches through the Internet for the time taken to reach Ilsan using the Olympic Expressway. However, the user does not start from Gangnam now, but departs from Gangnam after 30 minutes. At this time, it is supposed that information is received through the Internet that an accident occurs in the vicinity of Seongsu Bridge and it will take about one hour, although it usually takes 30 minutes to reach Ilsan. However, since the user does not start at present but departs after 30 minutes, information about the time taken from Gangnam to Ilsan after 30 minutes is needed. However, the present real-time traffic situation information only provides current traffic situation information, and it does not provide information on how the traffic situation changes on the Olympic road after 30 minutes and the running speed changes. Therefore, based on the information provided above, the user may use a riverside road or an alternative road instead of the Olympic road, but after 30 minutes, the traffic accident can be restored and the normal driving speed can be restored. Therefore, It may take an unexpectedly long time to drive to Ilsan using alternative roads. The existing traffic situation prediction system does not provide a solution to this problem, which is an inevitable result that the existing system originates from providing only information based on the past and present traffic conditions. However, in the case of an event (referred to as a future event in the present invention) that affects the traffic situation at a specific point in time such as a change in weather, a road construction in a characteristic road section, an event, Are not considered.

Therefore, according to the existing traffic situation prediction system, a means for predicting traffic, a means for allowing the system operator to make such a prediction, etc., are not provided by reflecting a future event, and means There is a demand for a traffic situation prediction system.

On the other hand, in the case of the driver, there are many plans to drive the vehicle to a specific area at a certain time during the day. That is, in many cases, it is often desired to know traffic information of a specific section within a specific time period from 1 hour to 24 hours. However, the existing traffic situation prediction system does not meet the needs of such users.

Disclosure of Invention Technical Problem [8] The present invention provides a traffic situation prediction system capable of predicting the effect of a future event occurring in the future on a traffic flow and providing the prediction to a user.

Another object of the present invention is to collect current traffic information, future event information, and weather information collected from an ITS center such as an external linking organization, for example, a local government, and transmit the traffic flow (for example, And provides a traffic situation prediction system that can predict the traffic situation and provide it to a user.

In order to achieve the above object, according to the present invention, there is provided a traffic information management system configured to receive and store current traffic situation information in various road sections including a predetermined road section in real time, Future event information, and unexpected situation information from an external traffic control center server configured to receive and store future event information and unexpected situation information including information of weather and weather data, A future traffic situation prediction server for predicting a future traffic situation in a predetermined road section up to 24 hours is provided. The server includes a database server configured to receive, process, and store current traffic situation information, future event information, weather information, and unexpected situation information for each road section from the external traffic control center server, A user interface module configured to control functions from a user's point of view and to display a result of the performance of the functions, and a controller configured to predict a traffic situation when a future event occurs in a predetermined road section and to provide a predicted result to the server operator A system for predicting a traffic situation, comprising: means for referring to data of a database server based on various traffic condition-related parameters input through the user interface module; Mid-long term traffic situation of And it characterized by including the traffic forecasting system configured to predict.

In one embodiment, the database server may also generate a vehicle speed pattern for each section based on link unit traffic information data for each road section received from the external traffic control center server, , The vehicle speed pattern may be modified by reflecting additional information related to the weather condition to generate and store speed pattern information corresponding to each situation.

In one embodiment, the user interface module includes a parameter input module for inputting parameter information required for predicting a traffic situation, and a parameter input module for estimating a traffic situation, An information providing module having a function of displaying the future traffic situation result, adjusting the time according to the user input and displaying the result, and allowing the user to analyze the traffic situation based on the result.

In one embodiment, the traffic situation prediction system uses speed pattern information stored in a DB server based on per-link speed information, unexpected situation information, construction information, event information, and weather information collected in real time, And a prediction module configured to predict a long-term traffic situation.

In one embodiment of the present invention, the traffic situation prediction system includes an analysis module for allowing the system operator to simulate the traffic situation prediction by adjusting the parameter used in the prediction in relation to the predicted future traffic situation through the prediction module Module. ≪ / RTI >

According to the present invention, a traffic situation in a specific road section from now until, for example, 24 hours is considered in consideration of future traffic information that has not been used in the past, such as event information on a specific road section, road construction information, weather data, Predicted and provided to the user.

1 is a diagram schematically illustrating a configuration of a traffic situation prediction server according to an embodiment of the present invention.
2 to 7 are views for explaining a vehicle speed prediction method using a prediction algorithm that can be used in predicting a vehicle speed in a traffic situation prediction server of the present invention.
FIGS. 8 to 12 are views showing exemplary user interfaces provided for each step when the traffic situation prediction server of the present invention is actually implemented.

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In the following description, the embodiments are merely exemplary and are not intended to limit the scope of the present invention. For example, each component of the system described below may be configured in hardware, software, or a combination thereof. In addition, it is explained that the traffic situation is predicted in cooperation with the local government ITS center, but it is not necessarily limited to the local ITS center. In addition, although it is described that the traffic situation prediction server is separate from the local autonomous ITS center, it may be integrated into the server of the local autonomous ITS center, which is also within the scope of the present invention.

1. System Configuration

As shown in FIG. 1, a traffic situation prediction server according to an embodiment of the present invention is a server configured to receive predetermined data from a local government ITS center server 100, process the data, and predict a traffic situation. The traffic condition prediction server includes a DB server (DB), a traffic situation prediction system (200), and a user interface module (300).

The DB server DB receives not only the current traffic information data from the external linking organization, in this embodiment, the municipal ITS center server 100 but also the traffic information such as future events such as weather data, construction information, event information, And a server for processing and storing data for use in the system of the present invention. That is, the DB server is configured to receive and store data required for prediction such as per-link speed data, traffic volume data, unexpected information, construction information, event information, and weather information collected in real time from the local government ITS center. In addition, the DB server stores speed pattern information. That is, although not shown, the DB server is provided with a pattern generation module. The pattern generating module is configured to generate a pattern of vehicle speeds per link (section) based on traffic situation information data provided from the outside. The pattern generation module generates normal speed patterns based on the traffic information data on the specific link provided from the ITS center server 100 and generates additional information related to day of week, holiday, event, construction, And corrects the speed pattern to generate real time speed pattern information corresponding to each situation, and the speed pattern information is stored in the DB server.

On the other hand, in the local government ITS center, current traffic situation information of an accelerated road such as a riverside north road or an Olympic highway is received in real time. In addition to this information, construction information, event information, weather data And event information such as a traffic accident is also received. In the system of the present invention, not only the current traffic information but also the future event information and the sudden situation information are received in real time, processed and processed to estimate the future traffic situation, for example, from 1 hour to 24 hours from the present, To the system operator (and ultimately the end user).

The user interface module 300 is a means for controlling functions performed in the system of the present invention from a user's point of view and expressing the result of performing the functions. The user interface module 300 receives parameters required for short- , And transmits the predicted traffic condition prediction system 200 to the user, and displays the predicted result to the user. The user interface, for example, as shown in FIGS. 8 to 12, is provided to the system operator.

Specifically, the parameter input module 310 is a module for inputting parameter information necessary for predicting a traffic situation such as a specific date and time to be predicted. The information providing module 320 checks and analyzes the traffic condition prediction execution result As a means for enabling the traffic situation prediction system 200 to display the future traffic situation result predicted by the traffic situation prediction system 200, time is adjusted according to user input to express the result, It is a function that allows you to do.

The user traffic situation prediction system 200 is a system for predicting a traffic situation when a future event occurs and transmitting a predicted result to a user. The system estimates a traffic condition based on various parameters input through the parameter input module 310, Short-term, medium / short-term traffic conditions are analyzed and predicted by referring to data of the DB server collected and stored in real time from the local autonomous ITS center server 100.

Specifically, the short-term prediction module 210 performs a predetermined prediction algorithm (described later) based on the current time based on the per-link speed information, the unexpected situation information, the construction information, the event information, ) To predict the traffic situation. At this time, the traffic situation is predicted using the speed pattern information stored in the DB server.

The medium / long term prediction module 220 is configured to predict a link-specific speed from 1 hour to 23 hours on the basis of the current time. The medium / long term prediction module 220 estimates the speed of each link based on a day, a holiday, an event, Traffic information is predicted using pattern information.

The analysis module 230 is a module for allowing a system operator to simulate a traffic situation prediction by adjusting parameters used for prediction in order to improve the accuracy and reliability of traffic situation prediction and is used for a prediction algorithm used in the prediction module And it is also possible to input data usage of generated pattern information according to day of the week, holiday, event, construction, unexpected situation, weather condition. Also, the reference data used in the prediction is provided to the operator so that the operator can analyze the prediction result.

The output module 240 is a module for providing the predicted traffic condition result performed by the prediction module and the analysis module to the user in the form of a graph or the like.

2. Vehicle speed prediction method (prediction algorithm)

Hereinafter, concrete embodiments of prediction algorithms available to the prediction modules 210 and 220 will be described in detail. The following prediction algorithm is only one example, and the present invention is not limited to the following prediction algorithm. That is, various prediction algorithms can be applied to the present invention, and it is possible to predict a future traffic situation (1 hour to 24 hours from now) through such an algorithm, and to provide, for example, It is to be understood that the present invention is characterized in that a traffic situation prediction server having the same configuration is constructed and that the present invention is not limited to the prediction algorithm itself.

First, a method of predicting future traffic information on the specific link using the vehicle speed acquired from the specific link to be predicted will be described.

FIG. 2 illustrates a method of predicting a vehicle speed using a prediction algorithm that can be used in the present invention. In this method, a road is divided into a plurality of links according to a traveling direction of a vehicle flow, Fig.

The traffic flow flows from the upstream side to the downstream side of the road, and the road includes a plurality of links. For example, the link 13 may be defined by a road space existing between two adjacent points 33, Sensors 21 to 25 are provided on each link. Each sensor is configured to measure the vehicle speed at each link. At this time, the vehicle speed may be an average value of all kinds of vehicle speeds measured at each link.

3 is a graph showing time-wise measured values 202 of the vehicle speed measured from the past to the present in the first link 13 of Fig.

3, the x-axis represents the flow of time, and the y-axis represents the vehicle speed measured at the first link 13 (0 represents the current). In the prediction algorithm, the time interval for measuring and providing the first vehicle speed on the first link 13 may be performed in units of a predetermined unit time (U). U may be a specific value set by the user and may have a value of 1 minute, 5 minutes, 10 minutes, etc. (5 minutes in the illustrated example). In the illustrated example, it is assumed that the specific time interval A0, that is, the interval from the near past to the present is the time interval from the past 20 minutes to the present. In FIG. 3, one set of five measured values is a first time-series value of the vehicle speed obtained at intervals of 5 minutes, and a graph 210 is a graph obtained by interpolating the time-series values.

4 is a graph showing the history of past velocity data obtained from the first link 13 in Fig. 2

Specifically, only N (N = 5) portions which are determined to be most similar to the traffic speed pattern in the specific time zone A0 shown in FIG. 3 are shown separately.

The following steps can be taken to derive the graph of FIG.

First, a step of generating a population to be compared with the traffic information (= traffic data) acquired in the specific time zone A0 shown in FIG. 2 may be performed. The traffic information relates to the speed of the vehicle and may be, for example, the first time series value comprising the vehicle speed measurement value 202 as shown in FIG.

The term " non-traffic information " used herein may mean information related to an external environment that is not related to the speed of the vehicle, such as a day of the week, a weather condition, a road construction event, The term " population " refers to a value obtained in the past for the first link 13, for comparison with the traffic data relating to the first link 13 acquired in the specific time period A0 shown in FIG. 3 (E.g., a total of one month's worth of data) extracted from all the traffic data (for example, data for the past 40 years). If the size of the population is too large, there is a disadvantage that the computing power required for the comparison becomes too large. If the size of the population is too small, there is a disadvantage that the comparison data required for prediction is not sufficient. Therefore, it is necessary to properly define the population. In the present invention, the following population extraction method can be used.

For example, the following two methods can be used to define the population.

First, the past time period of the same day as the day of the week corresponding to the specific time zone (A0) is K

You can choose to make a population.

Secondly, non-traffic information (e.g., non-traffic information similar to rainfall in the past) from non-traffic information related to the specific time zone A0 (e.g., non-traffic information similar to the specific time zone A0)

And the number of M times in the past that have the information about the past time interval). For example, if the rainfall has been falling in the specific time period A0, the past M time periods that have been rain in the first link may be selected and added to the population. The past M time periods are classified into various time periods, for example, Monday, Tuesday, and Friday

It can be one hour.

Therefore, the mother interval may include a total of K + M time periods. K and M are natural numbers from 1 to M and may be a value determined by a user. At this time, because the non-traffic information called 'day of the week' has already been used in the first method, the non-traffic information mentioned in the second method may be information excluding the 'day of the week' mentioned in the first step.

Next, among the above defined populations, it is possible to perform the step of finding N past time periods that match the first time series value 202 obtained during the specific time period A0 defined in FIG. The N may be a predetermined value by the user, and the meaning of 'matched' indicates the meaning of 'having a value included in a range of predetermined values.'

In the above method of the present invention, it is assumed that K and M each have a value of " 15 ", so that the population can be composed of 30 time periods. In addition, it is assumed that the current non-traffic information (event) is 'Rainy Friday', and N is '5'. Non-traffic information corresponding to past time periods PA1 to PA5, which are part of the population shown in FIG. 4, may be presented as shown in FIG. 5A. It is assumed that the lengths of the past time periods (PA1 to PA5) are the same as the specific time period (A0).

3 and FIG. 4 are compared with each other. FIG. 3 shows the history data measured for a short period of time, for example, 20 minutes before the present time. FIG. 3 shows a part of the selected population, will be.

The vehicle speed values measured during the past time periods PA1 to PA5 in FIG. 4 indicate values matched to the first time series value 202. Hereinafter, the vehicle speed values 100, 110, 120, 130, May be referred to as a 'first past time series value'.

The graphs indicated by reference numerals 102, 112, 122, 132, and 142 in FIG. 4 correspond to the vehicle speed measured and stored immediately after the five past time periods PA1 to PA5. The set of vehicle speeds represented by the graphs 102, 112, 122, 132, and 142 herein may be referred to as a 'following vehicle speed set'.

FIG. 5A is a table showing non-traffic information corresponding to each section extracted in FIG. Each column represents the past time interval of each section in FIG.

Hereinafter, a method of predicting future vehicle speed on the first link 13 shown in Fig. 1 will be described with reference to Figs. 1 to 4A together.

The vehicle speed prediction method using the prediction algorithm used in the present invention may include the following steps.

Referring to FIGS. 2 and 3, the prediction method includes a first time series value (measured value of vehicle speed) 202 relating to the vehicle speed at a specific time interval A0 obtained from the first link 13 (Step S10).

Next, the method further comprises the steps of: comparing N (ex: N = 5) first past time series values (100, 110, 120, 130, 140) closest to the acquired first time series value (Step S20) of finding a population (see FIG. 3) formed from the data history.

Then, the method according to the first embodiment of the present invention finds N = 5 first past time series values (100, 110, 120, 130, 140) For each of the past time periods (PA1 to PA5), for each of the time points (101, 111, 121, 131, 141) determined from immediately after the past time periods (PA1 to PA5) (S30) of the subsequent vehicle speed set (102, 112, 122, 132, 142) of the vehicle.

In the method, it is assumed that the predetermined time is one hour, and the subsequent vehicle speed set (102, 112, 122, 132, 142) is made up of a series of speed values, ~ PA5) means a set of vehicle speed values within one hour from immediately after.

The method then obtains each subsequent vehicle speed set 102, 112, 122, 132, 142 of the five past time periods PA1 through PA5 and then sets the subsequent vehicle speed set 102, 112, 122, 132, and 142. In this case, In this case, the average value may use five weight values determined by the distance between the first time series value 202 and each of the first past time series values 100, 110, 120, 130, and 140.

The average value calculation process between the following vehicle vehicle speed sets may be as follows.

≪ First step (S41) >

Create wf (1) to wf (5) through the following equations.

a value obtained by calculating the distance between wf (x) = [first past time series value, first time series value]

X may be a natural number of any one of the following numbers 1 to 5

the difference in distance between wf (1) = [first first past time series value (100), first time series value (202)

the difference in the distance between wf (2) = [the first first past time series value 110, the first time series value 202]

the difference in the distance between wf (3) = [the first first past time series value 120, the first time series value 202]

wf (4) = [the first first past time series value 130, the first time series value 202]

the difference in the distance between wf (5) = [fifth first past time series value 140, first time series value 202]

At this time, the smaller the magnitude of the absolute value of wf (x), the higher the weight for the first past time series value.

≪ Second Step (S42) >

Using the predetermined weight wb (x) corresponding to wf (x), a set of weights is obtained as follows.

First follower vehicle speed set 102 * wb (1) = first weight set

Second subsequent vehicle speed set 112 * wb (2) = second weight set

Third follower vehicle speed set 122 * wb (3) = third weight set

Fourth forth vehicle speed set 132 * wb (4) = fourth weight set

Fifth successive vehicle speed set 142 * wb (5) = Fifth weight set

≪ Third Step (S43) >

The average value between successive vehicle speed sets is equal to the average value of the weight set, e.g.

Average value between successive vehicle speed sets = (first weight set + second weight set + third weight set + fourth weight set + fifth weight set) / 5

In this example, the average value data between the successive vehicle speed sets can be regarded as the predicted future vehicle speed from the present to one hour on the first link 13.

Hereinafter, a method for predicting future traffic information on the specific link using the vehicle speed acquired from two links including a specific link to be predicted will be described.

The graphs 221, 231 and 241 shown in Fig. 6 are graphs showing the vehicle speeds from the near past to the present measured at the first link 13, the second link 12 and the third link 14 in Fig. (220, 230, 240) according to time.

The meaning of the x and y axes in Fig. 6 is as shown in Fig. 3, and the unit of time (U) of the specific time interval (A1) is defined as the unit of time (U) of the specific time interval . The graphs 221, 231, and 241 shown in FIG. 6 are time series values (for example, time series values) about the vehicle speed during the past past time (specific time period A1) obtained from the links 13, 12, (Measured values of vehicle speed) (220, 230, 240).

FIG. 7 is a graph showing the history of the velocity data obtained according to the condition of the population among the historical data obtained for the past several years for each link shown in FIG. 2. FIG.

The following steps can be taken to derive the graph of FIG.

First, a step of generating a population for comparing with the traffic information (= traffic data) acquired in the specific time interval A1 shown in FIG. 6 may be performed. Herein, the 'population' refers to the sum of the traffic data obtained for the first link 13, the second link 12, and the third link 14 obtained in the specific time period A1, (For example, a total of one month) of the past traffic data extracted from all traffic data (for example, data for the past 40 years) acquired in the past for the first link 13, the second link 12 and the third link 14 Amount of data).

For example, the following two methods can be used to define the population.

First, the past time period of the same day as the day corresponding to the specific time zone (A1) can be selected as K population.

Secondly, non-traffic information (ex: information on whether rain has occurred in the past on the first link) similar to non-traffic information (ex: rain on a specific time zone A1) related to the specific time zone A1, The population can be defined by further including M of the past time periods having. For example, if the rain has been falling in the specific time zone A1, M past time zones that have been rain in the first link may be selected and added to the population. The past M time periods may be time periods belonging to various days of the week such as Monday, Tuesday, Friday, and the like.

Therefore, the mother interval may include a total of K + M time periods. K and M are natural numbers from 1 to M and may be a value determined by a user. At this time, because the non-traffic information called 'day of the week' is already used in the first method, the non-traffic information mentioned in the second method may be the information excluding the information on the day of the week mentioned in the first step.

Next, in the above defined population, the second time series value 220, the third time series value 230, and the fourth time series value 240 acquired during the specific time interval A1 defined in FIG. 5 are simultaneously matched Lt; RTI ID = 0.0 > N < / RTI > past time periods. The N may be a predetermined value by the user. In the second embodiment of the present invention, it is assumed that K and M each have a value of " 15 ", so that the population may include 30 time periods. It is also assumed that the current non-traffic information (event) is 'Rainy Friday', and N is '3'. Non-traffic information corresponding to past time periods PB1 to PB5, which are part of the population shown in FIG. 6, may be as shown in FIG. 4B. It can be assumed that the length of each of the past time periods PB1 to PB5 is the same as the length of the specific time interval A1 shown in FIG.

The first graph 300 of FIG. 7 is a past data history for the second link 12, the second graph 400 is the first link 13, the third graph 500 is the third link 14, Lt; / RTI > Also, the past time interval (PB1 ~ PB5) refers to each past time interval extracted from past data history.

Reference numerals 310, 320, ..., 350 corresponding to the past time periods PB1 to PB5 represent vehicle speed values measured at the second link 12, reference numerals 410, 420, ..., 450 Represents the vehicle speed value measured at the first link 13. Reference numerals 510, 520,..., And 550 denote vehicle speed values measured at the third link 14.

Next, a subsequent vehicle speed set 312, 322, ..., 352 consisting of the values of the graph of predetermined time periods 311, 321, ..., 351 immediately after the past time periods PB1 to PB5, Can be defined. Next, a set of subsequent vehicle speeds 412, 422, ..., 452 consisting of the values of the graphs from immediately after the past time periods PB1 to PB5 to a predetermined time 411, 421, can do. Next, a set of subsequent vehicle speeds 512, 522, ..., 552 consisting of the values of the graphs from immediately after the past time period PB1 to PB5 to a predetermined time 511, 521, ..., can do. At this time, it is assumed that the length of the past time zone is 20 minutes, and the predetermined time is assumed to be one hour.

Hereinafter, with reference to Figs. 2, 5B, 6, and 7, a method for predicting future vehicle speed on the first link 13 will be described.

≪ Method of predicting future traffic information on the specific link using the vehicle speed acquired from two links including a specific link to be predicted -

The traffic speed on the first link 13 may be slower than the predetermined first value when a specific link, for example, a congestion occurs on the first link 13 of FIG. Whether or not the congestion occurs in the first link 13 can be predicted from the traffic data on the second link 12 located upstream of the first link.

In this case, it is assumed that the traffic speed on the first link 13 is lower than a predetermined first value, and therefore the traffic data on the second link 12, which is located upstream of the first link 13, .

The traffic flow forecasting method according to the present embodiment calculates the second time series value 220 and the third time series value 230 in the specific time interval A1 with respect to the first link 13 and the second link 12, (S110). ≪ / RTI >

Thereafter, in the past data 300 and 400 of the first link 13 and the second link 12, the acquired second time series value 220 and the third time series value 230 are simultaneously matched (S 120) of finding N pairs consisting of a second past time series value for the first link (13) and a third past time series value for the second link (12).

In this case, it is assumed that N is 3, and the three pairs of past time series values (320, 420, 330, 430, 350, 450) (PB2), the third past time period (PB3), and the fifth past time period (PB5). After finding the three pairs of the past time series values, it is possible to determine whether or not the subsequent time series of the first link 13 in the one time period 421, 431, and 451 from immediately after the past time periods PB2, PB3, Vehicle set 422, 432, and 452 is obtained. Calculating the average value between the following vehicle speed sets 422, 432 and 452 on the first link 13 in order to predict the vehicle speed for the next one hour from the present on the first link 13 Wherein the average value in the first link 13 can be calculated using three weights determined by the distance between the first time series value 220 and the first past time series values 420, have.

The average value calculation process between the following vehicle vehicle speed sets may be as follows.

<First step>

Create wf (1) to wf (3) through the following equations.

the distance between wf (x) = [second past time series value, second time series value]

X may be a natural number of 1 to 3

the difference in the distance between wf (1) = [first second past time series value 420, second time series value 220]

the difference in the distance between wf (2) = [second second past time series value 430, second time series value 220]

the difference in the distance between wf (3) = [third second historical time series value 450, second time series value 220]

At this time, the smaller the magnitude of the absolute value of wf (x), the higher the weight of the second past time series value can be.

<Second Step>

Using the predetermined weight wb (x) corresponding to wf (x), a set of weights is obtained as follows.

First follower vehicle speed set 422 * wb (1) = first weight set &lt; RTI ID = 0.0 &gt;

Second subsequent vehicle speed set 432 * wb (2) = second weight set

Third follower vehicle speed set 452 * wb (3) = third weight set

<Third Step>

The average value between the successive vehicle speed sets is equal to the average value of the weight set, and applies to the embodiment as follows.

Average value between successive vehicle speed sets = (first weight set + second weight set + third weight set) / 3

&Lt; Method of predicting future traffic information on the specific link using the vehicle speed acquired from two links including a specific link to be predicted -

1, the traffic speed on the first link 13 may be faster than the predetermined first value. Whether or not the traffic congestion will occur in the first link 13 when the traffic flow on the first link 13 is smooth can be predicted from the traffic data on the third link 14 existing downstream of the first link 13 have.

The traffic flow predicting method according to this case may include the same steps as the case 1 above. However, in this case, the first link 13 to be predicted and the third link 13 to be predicted in this case are used in the case 1. However, in the case 1, the first link 13 to be predicted and the second link 12, 14) is used.

In the above examples, two adjacent links are used, but the two links may not be adjacent to each other. For example, although the first link 13 and the second link 12 located upstream of the first link 13 are used in the above case 1, the first link 13 and the link 11 located further upstream may be used. In this case, the first link 13 and the third link 14 downstream of the first link 13 are used. Alternatively, the first link 13 and the link 15 located further downstream may be used.

&Lt; Case 3: Method of predicting future traffic information on the specific link using the vehicle speed acquired from three or more links including the specific link to be predicted &

In Case 2, traffic information previously predicted on the first link 13 is predicted in such a manner that the first link 13 to be predicted and one other link are used.

Alternatively, in case 3, the first link 13 to be predicted and two or more different links may be used. That is, Case 3 is an example of extending the concept of Case 2. For example, when traffic congestion occurs on the first link 13, traffic information obtained from the first link 13 and the link 11 and the link 12 can be used together. Alternatively, when the traffic flow is smooth on the first link 13, the first link 13 and the traffic information obtained from the link 15 and the link 14 can be used together.

According to the present case, the average value data between the successive vehicle speed sets may be the predicted future vehicle speed from the present to the first link 13 up to one hour.

In addition, predicting future vehicle speed using a plurality of links may be more accurate than predicting future vehicle speed using a single link.

3. Example

Hereinafter, a specific example of predicting a future traffic situation through the system of the present invention will be described according to an algorithm having the above-described configuration.

(1) Login process - User interface module

First, when the system of the present invention is connected to the system through a communication network, an interface such as that shown in FIG. 8A is provided, and an interface such as that shown in FIG. 8B is provided when a predetermined login process is performed. ①), schedule information (②) are provided.

(2) Short-term prediction - Traffic situation prediction system

When the short-term prediction button (1) provided through the user interface module 300 is selected, a parameter input menu (2) is provided (FIG. 9A).

As shown in FIG. 9B, parameters required for short-term prediction are provided, such as analysis interval selection, current time, k value, time set, historical data start and historical data finish. .

The current time, historical data start, and historical data finish can be specified as dates through the calendar, or as text separated by a '/' (eg 2014/09/01). In the case of time, it can be input through a clock or text.

The k value and the time set are entered as numbers. In the present invention, 'k value' means to find k traffic situations which are the most similar to the current traffic situation at the time of the short-term prediction analysis (when the algorithm is executed), and the time set means input to the time set based on the current time It means to process one set of continuous data by one set (the higher the input value, the longer the calculation time).

When the day of the week pattern is checked, it means that only the same day data is used during the past data search period.

10A, when the short-term prediction is performed after inputting the parameters according to the above procedure, all the links constituting the selected road section are outputted as a continuous link speed (see (1) &Lt; / RTI &gt; The time control bar (③) can be used to check the traffic conditions (predicted speed) for each link over time. By placing the mouse cursor in a specific section as shown in (2), detailed information about the link section can be confirmed. In addition, when a specific link is selected, the actual speed and the predicted speed in the link section are compared and displayed. The data used for the short-term prediction is displayed as shown in FIG. 10B, and the displayed data can be exported to the outside.

(3) Short-term Predictive Analysis - Analysis Module

The short-term prediction described above can be said to be a prediction based on an ideal value when performing a specific algorithm. However, since the prediction result may vary depending on the value input by the user, the system operator may change the values of the parameters and input the values, and analyze the short-term prediction results to improve the accuracy and reliability of the predicted traffic situation have.

When the analysis section is selected and the short-term prediction detailed analysis is selected, a parameter input menu required for this analysis is presented (FIG. 11A). The types of parameters are the same as in (2) above, and the system operator can input values with different values.

When the detailed analysis is completed, the IDs of all the links in the selected section are displayed as shown in FIG. 11B. When a specific link is selected, for example, a prediction result of 24 hours is displayed. In the example shown, the actual values for the selected link on the selected date and the predicted results after 15, 30, 45 and 60 minutes are shown. When the operator selects a predetermined predetermined point (e.g., 4), all the k pieces of data referred to for calculating the predicted value of the selected point are displayed as a graph, as shown in Fig. 11C. The displayed graph shows the traffic conditions of the referenced dataset by displaying one hour before and after the reference dataset (5-minute unit, time set parameter).

(4) Long term prediction - medium / long term prediction module

12A, when the analysis section and long term prediction are selected, a parameter input menu required for long term prediction such as short-term prediction is displayed (FIG. 12B). When long term prediction is performed after inputting all the parameters, the result as shown in Fig. 12C is displayed. Although the prediction is substantially the same as the short-term prediction, the prediction result for 23 hours from 1 hour is displayed based on the local time.

Claims (5)

The information processing apparatus is configured to receive and store current traffic situation information in various road sections including a predetermined road section in real time, and to receive future information including future information on construction information, event information, and weather data, Future event information, and unexpected situation information from an external traffic control center server configured to receive and store information and unexpected situation information from the external traffic control center server, and transmits the current traffic situation information, future event information, As a future traffic situation prediction server for predicting the situation,
A database server configured to receive, process and store current traffic situation information, future event information, weather information, and unexpected situation information from the external traffic control center server by road section,
A user interface module configured to control functions performed by the future traffic situation prediction server from a user's point of view and to display results of performing functions;
A traffic situation prediction system for predicting a traffic situation when a future event occurs in a predetermined road section and providing a predicted result to the server operator, the system comprising: A traffic situation prediction system configured to analyze and predict a short-term traffic situation from the present time to one hour and a long-term traffic condition from one hour to 23 hours with reference to data of the database server,
A traffic forecasting server for estimating a future traffic situation. The database server according to claim 1, wherein the database server further generates a vehicle speed pattern for each section based on link unit traffic information data for each road section received from the external traffic control center server, And generates and stores speed pattern information corresponding to each situation by modifying the vehicle speed pattern by reflecting additional information related to the situation. [Claim 4] The method according to claim 2, wherein the user interface module comprises: a parameter input module for inputting parameter information necessary for predicting a traffic situation; And an information providing module having a function of displaying a result of the situation, adjusting the time according to user input and displaying the result, and allowing the user to analyze the traffic situation based on the result, Situation prediction server. The traffic situation prediction system according to any one of claims 1 to 3, wherein the traffic situation prediction system includes speed pattern information stored in the DB server based on per-link speed information, unexpected situation information, construction information, event information, And a prediction module configured to estimate a short-term traffic condition and a long-term traffic condition using the prediction module. The traffic situation prediction system according to claim 4, wherein the traffic situation prediction system includes an analysis module for enabling a system operator to simulate a traffic situation prediction by adjusting a parameter used for the prediction in relation to the future traffic condition predicted through the prediction module Wherein the server further comprises:

Images