KR20020017535A - A method estimating the section-velocity for the advance direction - Google Patents

Abstract

PURPOSE: A method for predicting a sectional speed according to driving directions of a vehicle is provided to supply useful information to a driver by calculating correctly a sectional speed of a driving vehicle. CONSTITUTION: An information analysis process is performed to analyze various information of a probe vehicle received from a vehicle information relay portion(S10). A sectional speed calculation process is performed to calculate a sectional speed of a vehicle by using various data(S20). A driving direction of the vehicle is determined by using speed information about two roads(S30). A sectional speed of the driving direction is calculated after the driving direction of the vehicle is determined(S40). A current sectional speed prediction process is performed to predict the current speed of the vehicle(S50). The obtained traffic information is provided to a driver(S60).

Description

TECHNICAL FIELD [0001] The present invention relates to a method for real-

The present invention relates to traffic information collection and provision for a road, and more particularly, to a vehicle speed detection method and a vehicle speed detection method for detecting an accurate section speed for each direction in which a running vehicle is traveling, The present invention relates to a method for real-time prediction of an interval speed of a vehicle according to a traveling direction of the vehicle.

There are two types of conventional traffic information collection systems provided are the section detection method and the point detection method.

First, in the section detection method, a probe vehicle (PROBE CAR) travels on a road, and transmits information such as a travel time and a travel distance to a road section to a traffic information center using a wireless communication or a wired communication network, Is a method of calculating traffic information on each road using this information.

The point detection method is a method in which a point detector (loop detector, infrared camera, etc.) is installed on the road, the traffic volume occupancy rate obtained from the device is transmitted to the traffic information center, and traffic information on each road is calculated in the traffic information center to be.

Both of these methods can be applied to one road section (a road between an intersection and an intersection, a road between a road junction and a road junction, a road between an intersection and a road junction (hereinafter referred to as a 'link') The speed information of the link is calculated regardless of the direction in which the vehicle traveling the link travels at the preceding intersection or the road joining point.

Therefore, such information has an average meaning in each direction of travel. Therefore, in the case of running straight ahead, speed information lower than actual speed is given to a vehicle traveling on the left or other methods (right turn, U turn, P turn, etc.) There is a problem that information is provided.

As a result, since the conventional method does not consider the traveling direction of the vehicle at all, accurate speed information can not be provided to the traveling vehicle, and speed information considering the traveling direction can not be provided in terms of the traffic signal control, Lt; RTI ID = 0.0 > controllable < / RTI >

In order to calculate the traffic information for each direction by the above-mentioned point detection method, it is necessary to install the traffic information at an intersection or a road junction point. In this case, two point detectors must be installed in order to consider forward and reverse travel for each link, In this case, there is a problem in that it is costly and there is a problem in calculating the accurate speed of the link.

Further, in order to calculate the traffic information for each traveling direction by the section detection method, there are many problems in providing real-time speed information due to an increase in the system load due to the troublesomeness of tracking the travel route of each probe vehicle and the calculation amount problem have.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and it is an object of the present invention to provide a vehicle speed control system and a vehicle speed control method, To provide a real-time prediction method of the section speed for each traveling direction of the vehicle.

It is another object of the present invention to provide a vehicle speed control apparatus and a vehicle speed control method for simultaneously transmitting speed information of two consecutive links of a vehicle, And to provide a real-time prediction method of the star-section velocity.

In order to achieve the above object, the present invention provides a vehicle information analyzing method for analyzing various information of a probe vehicle received from a vehicle information relaying unit, calculating a section speed of the vehicle using various information such as a link travel distance and a link stop time And a speed information unit for determining a current traveling direction of the vehicle through speed information for two consecutive roads at a speed of the past road and a previous road at the time of generating the speed information unit through the vehicle information A step of determining a traveling direction of the vehicle and a step of calculating a traveling speed of each of the vehicles based on the traveling speed of the vehicle when the traveling direction of the vehicle is determined, And a traffic information providing step of providing the traffic information to the driver of the vehicle, Standing, is continuous to the vehicle velocity information of the roads in the current interval in which the measurement is required by the vehicle speed information of the road section is characterized in that connected after the prediction and by further comprising a current section rate prediction step of calculating.

1 is a configuration diagram of a vehicle speed detection system on a road to which the present invention is applied;

FIG. 2 is a block diagram illustrating an example of the present invention,

3 is a diagram illustrating a neural network structure for predicting real-time velocity information according to the present invention.

4 is a flowchart illustrating a method of real-time prediction of a section speed according to the traveling direction of the vehicle according to the present invention.

Hereinafter, the present invention will be described in more detail with reference to the accompanying drawings.

1 is a configuration diagram of a vehicle speed detection system on a road to which the present invention is applied.

As shown in the figure, 10a, 10b, ... are a plurality of probe vehicles running on the road, and various kinds of traffic information collected from the vehicle information collecting means 20a, 20b, , 30b, ...) wirelessly.

The vehicle 10 is mounted on the probe vehicles 10a, 10b, ... so as to measure the distance, time, or stopping time of the vehicle, Various traffic information is detected.

30a and 30b are vehicle information relay means and are provided at predetermined positions such as an intersection point and a road confluence point (hereinafter referred to as a node (NODE)) and are transmitted from the probe vehicles 10a, 10b, The traffic information of the vehicle is transmitted to the traffic information control center 40 or directly to the traffic light control processor (not shown).

Reference numeral 40 denotes a traffic information control center which receives various kinds of vehicle information from a plurality of the vehicle information relay means 30a, 30b,... By wire or wirelessly, calculates required traffic information, .

When the probe vehicle 10b enters a predetermined reception detection distance of the vehicle information relaying means 30b as shown in the figure, the vehicle information relaying means 30b transmits the probe vehicle 10b and various kinds of information And transmits the received information to the traffic information control center 40.

As a result, when the vehicle equipped with the vehicle information collecting means 20a, 20b,... Passes through the plurality of vehicle information relay means 30a, 30b, The vehicle information relay means 30a, 30b,... Transmits the information to the traffic information control center 40 via the wireless or wired network again.

The vehicle information collection means mounted on the probe vehicle detects and generates the following information.

① Current node (C): The unique number of the node that has passed at the time when the probe vehicle generates the information unit.

(2) First previous node (P): the unique number of the node that has just passed before the probe vehicle passes the current node (C).

(3) Second previous node (PP): the unique number of the node that has just passed the probe vehicle past the first previous node (P).

④ First Previous Link (P-C): Information that distinguishes the past link from the first previous node (P) to the current node (C).

(5) First Previous Link (PP-P): Information that distinguishes the past link from the second previous node (PP) to the first previous node (P).

⑥ First previous link distance (P-C / S): Distance traveled from the first previous node (P) to the current node (C) (unit: m).

(7) Time required for first link (P-C / T): Time (unit: second) taken from the first previous node (P) to the current node (C).

(8) First previous link stop time (P-C / ST): sum of the times when the vehicle is stopped from the first previous node (P) to the current node (C) in seconds.

(9) Second previous link distance (PP-P / S): Distance traveled from the second previous node (PP) to the first previous node (P) in m.

⑩ Time required for the second previous link (PP-P / T): Time (unit: second) taken from the second previous node (PP) to the first previous node (P).

(11) Second previous link stop time (PP-P / ST): The sum of the times when the car is stopped while traveling from the second previous node (PP) to the first previous node (P).

⑫ Number of previous link stoppage (PP-P / SN) and current link stoppage (P-C / SN): Number of times when the vehicle stopped on the previous road (link)

⑬ History Index: Indicates the version of vehicle information collection means.

⑭ Message Index: Information that prevents re-registration of information.

⑮ CheckSum CRF (CheckSum CRF): Information for detecting the bit error of transmitted information.

Here, the first previous link distance PC / S and the first previous link time PC / T, the second previous link distance PP-P / S, and the second previous link time PP- / T are used to calculate vehicle speed information for the first previous link PC and the second previous link PP-P, respectively.

Therefore, the traffic information control center 40 calculates the section speed of the vehicle through the data received from the vehicle information relay means 30a, 30b, ....

At this time, the calculation method of the section speed is examined.

First, the vehicle speed (P-C / V) relative to the first previous link (P-C)

The first previous link speed (P-C / V) = [first previous link distance (P-C / S) / first previous link required time (P-C / T)] 3.6 (unit: Km /

The vehicle speed (PP-P / V) for the second previous link (PP-P)

(Second link distance PP-P / S) / the first previous link linkage time PP-P / T 3.6 (unit: Km / h) .

FIG. 2 is a road configuration diagram illustrating an example for explaining the present invention.

As shown in the figure, a plurality of nodes are formed at the intersections on the checkered road. Here, in accordance with the traveling direction of the probe vehicle 10, a second previous node PP, a first previous node P, C), and the vehicle information relay means 30a, 30b, ... are located at predetermined positions in the node PP, P, and C, respectively.

The traveling direction of the probe vehicle 10 is firstly advanced through the second previous node PP and then passed through the second previous link PP-P and is turned right at the second previous node P, Passes through the link (PC), and is now in a state of being straight at the node C. [ The current position of the vehicle 10 is the position that has passed through the current node C. [

Accordingly, the data transmitted from the vehicle information relay means 30a, 30b, ... located in the respective nodes C, P, PP through the current probe vehicle 10, (PC) and the second preceding link (PP-P). Of course, the vehicle speed information provided at this time is a case in which the vehicle passes right through the second preceding node PP to the first preceding node P and the vehicle passes the present node C, .

In this way, the traveling direction and the section speed of the vehicle are calculated. If the flow of traffic is smooth, the time zone passed by the first previous node P at the second previous node PP, And the time zone past the current node (C). However, if not, the speed information of the first previous node P in the second previous node PP indicates the speed information of the previous service unit time zone. In this case, the speed information of the current service unit time zone should be predicted (calculated) from the speed information of the previous time zone. This provides very accurate real-time speed information for the current road.

As a result, the speed information on the previous road is predicted based on the speed information of the previous time zone.

In the present invention, two methods are used to calculate the vehicle speed of the current road. One is using Kalman filter, and the other is using Neural Network.

First, we explain the Carmen method.

The Kalman scheme uses the following terms.

T: Number of speed information per link, created for servicing throughout the day. For example, if a 5 minute service is provided, 288 speed information is generated per link per day.

L: Provide rate service The total number of links you want to receive.

_{^{y i t, r, t =}} 1, ..., T, i = 1, ..., L, r = 1,2,3: link i t of the traveling direction information rate as measured on the time zone.

α ⁱ _{t, r is the} state variable of the link i at time t in the direction of travel, 1 × m vector,

ⁱ _{t, r is the} noise variable of the link i in the time direction of the time zone, n × 1 vector,

ⁱ _{t, r is the} noise variable for the direction of travel of the time zone of link i, g × 1 vector,

A ⁱ _{t, r} : 1 × m vector, B ⁱ _t : 1 × n vector,

C ⁱ _{t, r} : m × m vector, B ⁱ _t : m × g vector,

The following state space model is established for each link and direction for speed service according to the progress direction. The establishment of this model is done through statistical methods such as time series analysis.

<State Space Model>

y ⁱ _{t, r} = A ⁱ _{t, r} α ⁱ _{t, r} + B ⁱ _{t, r ^{i t, r, t = 1}} , ..., T, i = 1, ..., L, r = 1,2,3 ( measurement equation)

α ⁱ _{t, r} = C ⁱ _{t, r} α ⁱ _{t-1, r} + D ⁱ _{t, r ^{i t, r, t = 1}} , ..., T, i = 1, ..., L, r = 1,2,3 ( transfer equation)

here, { ⁱ _{t, r} }, { ⁱ _{t, r} } are independent of each other and are a probability process called white noise. The mean is all 0 and the covariance matrix is H ⁱ _{t, r} , Q ⁱ _{t, r,} respectively.

Therefore, when estimating the vehicle speed of the current road using the Kalman filter method, the speed can be recursively predicted using two equations, that is, a prediction equation and an update equation.

The least square estimator of the state variable α ⁱ _{t-1, r} is α ⁱ _{t-1, r} , and the covariance matrix of the errors α ⁱ _{t-1, r} - α ⁱ _t-1, ² P ⁱ _{t-1, r} , where P ⁱ _{t-1, r} is the known matrix, and the prediction equation is

<Prediction Equation>

α ⁱ _{t | t-1, r} = C ⁱ _{t, r} α ⁱ _{t-1, r} ,

ⁱ _{t | t-1, r} = A ⁱ _{t, r} α ⁱ _{t | t-1, r}

_{^{P i t | t-1,}} r = C i t, r P i t-1, r C i t, r '+ D i t, r Q i t, r D i t, r'

t = 1, ..., T, i = 1, ..., L, r = 1,2,3

here, ⁱ _{t | t-1, r} is the predicted value of the speed information at time t using data up to t-1.

In addition, the update equation uses the data up to t-1 to estimate the velocity information at time t ⁱ is the equation that provides the correction value for the speed information at time t when the speed information at time t is observed after providing _{t | t-1, r} .

<Update Equation>

_{^{P i t, r = P i}} t | t-1, r -P i t | t-1, r A i t, r '(F i t, r) -1 A i t, r P i t | t _{-1, r} ,

_{^{α i t, r = α i}} t | t-1, r + P i t | t-1, r A i t, r (F i t, r) -1 (y i t, r - A i t, _r ? ⁱ _{t | t-1, r} ),

_{^{F i t, r = A i}} t, r P i t | t-1, r A i t, r '+ B i t, r H i t, r B i t, r,

t = 1, ..., T, i = 1, ..., L, r = 1,2,3

Therefore, the velocity information at the (t + 1) th time point can be recursively predicted using the update equation and the prediction equation.

On the other hand, the second method for calculating the vehicle speed of the present road according to the present invention is a method using an algorithm known as a neural network.

The neural network is an effective analysis method for the optimization problems of complex and difficult models due to its self-learning ability. It uses the speed information of the offline database for each link and time zone, and inputs the speed information of the previous time zone and the speed information of the peripheral link And the predicted value is determined from the determined model by learning the neural network with the number of input neurons and one output neuron. The neural network is determined separately for each day of the week, link, and time zone. After midnight, it is learned based on the speed information of the day, and the newly learned model is used as a prediction model of the speed information of the same day after one week. When a road situation (construction, etc.) occurs or the road information is changed, it is a way to create a new neural network. FIG. 3 is a block diagram illustrating a neural network structure for predicting real-time velocity information according to the present invention.

The operation of the present invention configured as described above will be described below with reference to the flowchart of FIG.

4 is a flowchart illustrating a method of real-time prediction of a section speed according to the traveling direction of the vehicle according to the present invention.

As shown in the figure, the present invention can be applied to a vehicle information analyzing process (S10) for analyzing various information of the probe vehicle 10 received from the vehicle information relaying means 30, (S20) for calculating a section speed of the vehicle using various data, and a speed information calculating unit (S20) for calculating speed information on two consecutive roads at the time of generating the speed information unit through the vehicle information (S30) of determining a traveling direction of the vehicle traveling through the road, and a step of calculating a section speed calculation process for each traveling direction of the vehicle that calculates the section speed along the traveling direction when the traveling direction of the vehicle is determined (S50) of predicting and calculating the vehicle speed with respect to the current time of the previous link in the section speed calculation (S50) It consists of information to the traffic information providing process (S60) to give the driver of the vehicle.

The current section speed predicting step S50 of the vehicle includes a current section speed predicting step S51, a calculation method according to the current section speed predicting step S52, a Calmen filter method according to the calculation method selection (S53, S54) for judging whether or not the method uses a neural network, and calculating the current section speed in real time by the two methods.

More specifically, the probe vehicle 10 equipped with the vehicle information collecting means 20 wirelessly transmits information about various vehicles to the vehicle information relaying means 30 installed at each node while driving the road. The information at this time is the current node C, the first previous node P, the second previous node PP, the first previous link PC, the first previous link PP-P, 1 previous link distance (PC / S), first previous link required time (PC / T), first previous link stop time (PC / ST) (PP-P / ST), the first previous link stop count (PP-P / SN), the current link stop count (PC / SN) A History Index, a Message Index, and a CheckSum CRF.

Therefore, the vehicle information relay means 30 receiving this information transmits this information to the traffic information control center 40 or directly to the traffic light control processor (not shown). Here, the traffic information control center 40 analyzes various information of the probe vehicle 10 that has been transmitted (step S10).

Then, the speed of the link of the vehicle is calculated through the transmitted data. Here, the section speed for the vehicle is detected not only in the first previous link (P-C) where the current vehicle has passed but also in the speed for the second previous link (PP-P) (S20).

When the current position and the section of the probe vehicle 10, that is, the link speed, are determined as described above, the traveling direction of the vehicle can be determined. That is, the traveling direction of the vehicle on the second previous link PP-P, that is, whether the vehicle is a straight ahead vehicle or a right-turn vehicle, via the presently passing node C and the previous node P and the previous previous node PP , A left-turning vehicle, a U-turn vehicle, or the like.

When the traveling direction of the vehicle is determined in this way, the section speed of each vehicle in the traveling direction can be calculated (S40). Here, if the traffic flow is smooth, it belongs to the time zone of the first previous node P, the time zone of the first previous node P, and the time zone of the current node C, will be. However, if not, the velocity information of the temporal time should be predicted and calculated using the velocity information of the second previous link (PP-P).

Therefore, the traffic information control center 40 selects the Kalman filter method and the neural network method in order to calculate the speed of the vehicle on the current road. First, in the case of the Kalman filter method, the speed information for the section in which the vehicle is currently traveling is recursively calculated using the update equation and the prediction equation.

In the case of the neural network method, the predicted value is calculated from the determined model by learning from the neural network having the number of the input neurons and one output neuron using the speed information of the previous time zone and the speed information of the peripheral link as the input elements

In this way, the speed of the vehicle with respect to the road in the current time zone is calculated (S50).

The generated traffic information, that is, the section speed information for each traveling direction of the vehicle is again provided to the driver of the vehicle (S60). Therefore, drivers are provided with optimal traffic information according to the traveling direction of the vehicle reaching the destination.

As described above, according to the present invention, not only the section speed at the intersection or the road joining point but also the speed information about the roadway of the current on-going vehicle are provided in real time, And it is advantageous that the quality of traffic information provision is further improved.

Claims (3)

A vehicle information analyzing process for analyzing various information of the probe vehicle received from the vehicle information relaying unit and a vehicle speed information calculating process for calculating the section speed of the vehicle using various information such as link travel distance and link stop time And determining a traveling direction of the vehicle based on the speed information of two consecutive roads at a speed of the past road and the previous road at the time of generating the speed information unit through the vehicle information, If the traveling direction is determined, a traveling speed calculation process for each traveling direction of the vehicle that calculates the traveling speed of each vehicle according to the traveling direction, and a traffic information providing process for providing the driver of the vehicle with various traffic information corresponding to the calculated traveling speed A method for calculating a section speed for each traveling direction of a vehicle, And a current section speed prediction process of predicting and calculating the vehicle speed information of the road section after the vehicle speed information of the road in the current section requiring measurement is continuously connected in the section speed calculation, Real Time Prediction Method of Interval Velocity by Vehicle Direction. The method according to claim 1, Wherein the current section speed prediction method is a Kalman filter method that recursively calculates speed information for a road section in a current time zone using an update equation and a prediction equation. The method according to claim 1, The current section speed predicting method is a neural network method of calculating from a model determined by learning with a neural network having the number of input neurons and one output neuron using the speed information of the previous time zone and the speed information of the peripheral link as input elements A method for real - time prediction of a section speed of a vehicle in a traveling direction.