JP6807215B2 - Controls, estimation methods and programs - Google Patents

Description

The present invention relates to a technique for estimating the end position of a vehicle congestion.

Conventionally, there has been known a technique for measuring the traffic volume of a vehicle or the like by detecting the vehicle with a sensing device that detects the vehicle using ultrasonic waves or the like. Generally, the sensing device is installed in the inflow path where the vehicle flows into the intersection, and the traffic volume of the vehicle in the inflow path is measured.

The measured information such as the traffic volume of the inflow route is used for determining the presence or absence of congestion in the inflow route and estimating the end position of the congestion. Here, if the sensing devices are arranged at short intervals in the inflow path, the end position of the traffic jam can be accurately determined. On the other hand, since the sensing devices are relatively expensive, even if many sensing devices are installed, the sensing devices are generally installed at intervals of several hundred meters. Therefore, the estimation of the end position of the traffic jam is often inaccurate.

Patent Document 1 below discloses a technique for estimating the end position of a traffic jam in an inflow path. In the technique described in Patent Document 1, first, the degree of queuing spillover (the degree to which the queuing of the vehicle spills over at the sensing point) is obtained from the speed of the vehicle at the sensing point (paragraphs [0070] to [0075]. ], Fig. 4). Then, the end of the traffic jam of the vehicle is estimated from the queuing spread of the vehicle at each detection point (paragraphs [805] to [0086], FIG. 6A).

Japanese Unexamined Patent Publication No. 2016-133942

In the technique described in Patent Document 1, the end position of the traffic jam is estimated from the queuing spread degree of the vehicle at each sensing point, but this technique also has insufficient estimation accuracy.

In view of the above circumstances, an object of the present invention is to provide a technique capable of improving the accuracy of estimating the end position of a traffic jam.

In order to achieve the above object, the control device according to one embodiment of the present invention includes a control unit.
The control unit acquires the traveling information transmitted from the vehicle, acquires the sensing information obtained by detecting the vehicle, determines the end position of the traffic jam at the past time based on the traveling information, and determines the end position of the traffic jam. The correlation between the end position of the traffic jam at the past time and the sensed information at the same time as the past time is determined, and based on the correlation, the sensed information at the present time is used at the present time. The end position of the traffic jam is estimated.

As a result, the accuracy of estimating the end position of the traffic jam can be improved.

In the control device, the control unit may learn the correlation by a learning function.

In the control device, the control unit generates an estimation parameter based on the correlation, applies the estimation parameter to the sensing information at the current time, and determines the end position of the traffic jam at the current time. Control device to estimate.

As a result, the end position of the traffic jam can be estimated appropriately.

In the control device, the control unit applies the estimation parameter to the sensing information at the past time to estimate the end position of the congestion at the same time as the past time, thereby estimating the estimation parameter. The quality of the above-mentioned estimation parameters may be updated according to the evaluation result.

This allows the estimated parameters to be updated appropriately.

In the control device, the control unit may control the display time of the signal light based on the end position of the traffic jam at the estimated current time.

In this control device, the display time of the signal light can be controlled based on the accurate end position of the traffic jam, so that the traffic jam can be alleviated.

In the control device, the control unit may generate congestion information for transmission to the vehicle based on the end position of the congestion at the estimated current time.

In this control device, the traffic jam information can be generated based on the accurate end position of the traffic jam, so that the accurate traffic jam information can be generated.

The estimation method according to one embodiment of the present invention acquires travel information transmitted from a vehicle, acquires sensing information obtained by sensing the vehicle, and based on the travel information, causes a traffic jam at a past time. The end position is determined, the correlation between the end position of the traffic jam at the past time and the sensed information at the same time as the past time is determined, and the sensed information at the current time is determined based on the correlation. From, the end position of the traffic jam at the current time is estimated.

The estimation method according to one embodiment of the present invention is based on a step of acquiring travel information transmitted from a vehicle, a step of acquiring sensing information obtained by sensing the vehicle, and past travel information. Based on the step of determining the end position of the traffic jam at the time, the step of determining the correlation between the end position of the traffic jam at the past time, and the sensed information at the same time as the past time, and the correlation. Then, the computer is made to perform a step of estimating the end position of the traffic jam at the current time from the sensed information at the current time.

As described above, according to the present invention, it is possible to provide a technique capable of improving the accuracy of estimating the end position of a traffic jam.

It is a figure which shows the state when the traffic jam occurs with the intersection at the head. It is a figure which shows the control device which concerns on one Embodiment of this invention. It is a figure for demonstrating the basic idea of this invention. It is a figure which shows the process at the time of learning (determining) the correlation of the end position of a traffic jam and the sensed information in the past time. It is a figure which shows the process at the time of estimating the end position of a traffic jam based on the correlation at the present time.

Hereinafter, embodiments according to the present invention will be described with reference to the drawings.

<Configuration of control device 10 and configuration of each part>
FIG. 1 is a diagram showing a state when a traffic jam occurs, starting from an intersection. FIG. 2 is a diagram showing a control device 10 according to an embodiment of the present invention.

In the example shown in FIG. 1, a road extending in the north-south direction and a road extending in the east-west direction intersect at a right angle at an intersection. In the description of the present specification, the road for the vehicle 5 to flow into the intersection is referred to as an inflow path 1, and the road for the vehicle 5 to flow out from the intersection is referred to as an outflow path 2.

In FIG. 1, the road on the south side of the intersection is west of the central line 3, the road on the north side of the intersection is east of the central line 3, and the road on the east side of the intersection is on the south side of the central line 3. Of the roads and the roads on the west side of the intersection, the road on the north side of the central line 3 is the inflow road 1.

On the other hand, the roads on the south side of the intersection on the east side of the central line 3, the roads on the north side of the intersection on the west side of the central line 3, and the roads on the east side of the intersection on the north side of the central line 3. Of the roads on the west side of the intersection, the road on the south side of the central line 3 is the outflow road 2.

The intersection shown in FIG. 1 is a major intersection where traffic congestion is frequently caused, and traffic congestion occurs in the inflow path 1 on the south side of the intersection, starting from this intersection. In the inflow path 1 on the south side of the intersection, another intersection (signal light 20) may exist in the middle of the inflow path 1. When another intersection exists, the inflow path 1 that flows into the other intersection is also the inflow path 1 of the main intersection shown in FIG.

At the intersection, four signal lights 20 are provided corresponding to the four inflow paths 1. The vehicle 5 is on the left side (may be on the right side), and can turn right, go straight, or turn left at the intersection depending on the lighting state of the signal light 20.

The vehicle 5 is equipped with an in-vehicle device (not shown). This in-vehicle device communicates with GPS satellites (GPS: Global Positioning System), acquires time information, position information of the vehicle 5 at that time, and the like at predetermined intervals, and probe information (travel information). ).

Here, the in-vehicle device is not installed in all vehicles 5, but is installed only in some vehicles 5 (for example, about 10 to 20%). In the following description, when the vehicle 5 equipped with the in-vehicle device is particularly distinguished from all the vehicles 5, the vehicle 5 is referred to as a probe vehicle 5a (see FIG. 3).

Each of the four outflow paths 2 is provided with a road device 40 that wirelessly transmits and receives information to and from an in-vehicle device mounted on the vehicle 5. The road device 40 is, for example, a local communication device such as an optical beacon or a radio beacon, and the road device 40 is installed near an intersection in the outflow path 2. The road devices 40 may be installed not only in the outflow path 2 but also in the inflow path 1 on the south side of the intersection at predetermined intervals.

The road device 40 has a road device main body 41 and a communication device 42. The communication device 42 receives probe information from the vehicle-mounted device of the vehicle 5 that has entered the communication area, and also transmits traffic information (traffic jam information, traffic closure information, etc.) to the vehicle-mounted device. The road device main body 41 comprehensively controls each part of the road device 40 and stores information necessary for transmission / reception. Further, the road device main body 41 receives traffic information from the control device 10 by wire or wirelessly, and transmits probe information to the control device 10.

In the inflow path 1 on the south side of the intersection, three sensing devices 30 are installed along the inflow path 1. The first sensing device 30 is installed at a position 300 m from the stop line 4 of the inflow path 1, and the second sensing device 30 is installed at a position 500 m from the stop line 4. Further, the third sensing device 30 is installed at a position 1000 m from the stop line 4.

The position where the sensing device 30 is installed is not limited to 300 m, 500 m, and 1000 m from the stop line 4, and of course other values can be taken. Further, the number of sensing devices 30 is not limited to three, and may be one, two, or four or more.

The sensing device 30 has a sensing device main body 31 and a sensing device 32. The sensor 32 senses the vehicle 5 at the sensing point and acquires the sensing information by a method such as an ultrasonic method, an electromagnetic induction method, or an image acquisition method.

The sensed information includes information on the traffic volume of the vehicle 5, information on the occupancy rate of the vehicle 5, and information on the speed of the vehicle 5. The traffic volume information of the vehicle 5 is information indicating the number of vehicles 5 passing through the sensing point in a unit time (for example, 1 hour). The information on the occupancy rate of the vehicle 5 is information indicating the ratio of the time during which the vehicle 5 is detected at the sensing point to the unit time in a unit time (for example, 1 hour). The speed information of the vehicle 5 is information indicating the speed of the vehicle 5 passing through the sensing point.

The sensing device main body 31 comprehensively controls the sensing device 30 and stores the sensing information obtained by the sensing device 32. Further, the sensing device main body 31 transmits the sensing information to the control device 10 by wire or wirelessly.

The control device 10 is installed in a traffic control center that manages and controls the traffic of the vehicle 5 in a specific area. The control device 10 may be installed near the intersection, or may be incorporated in a traffic light including a signal light 20.

As shown in FIG. 2, the control device 10 includes a control unit 11, a storage unit 12, and a communication unit 13. The communication unit 13 communicates with the road device 40, the sensing device 30, and the signal light 20 by wire or wirelessly. Specifically, the communication unit 13 transmits traffic information to the road device 40 and receives probe information from the road device 40. Further, the communication unit 13 receives the sensing information from the sensing device 30. Further, the communication unit 13 transmits an instruction signal for controlling the display time (lighting time) of the signal light 20 to the signal light 20.

The storage unit 12 includes a volatile memory used as a work area of the control unit 11, various programs required for processing of the control unit 11, and a non-volatile memory for storing various information. The various programs may be read from a portable recording medium such as an optical disk or a semiconductor memory, or may be downloaded from a server device on a network.

The control unit 11 is composed of a CPU (Central Processing Unit) and the like. The control unit 11 executes various calculations based on various programs stored in the storage unit 12, and controls each unit of the control device 10 in an integrated manner. In the present embodiment, the control unit 11 mainly executes a process of learning (determining) the correlation between the end position of the traffic jam and the sensed information, or based on this correlation, from the sensed information. Executes the process of estimating the end position of the traffic jam. The specific processing of the control unit 11 will be described in detail below.

<Basic idea>
Next, the basic concept of the present invention will be described. FIG. 3 is a diagram for explaining the basic concept of the present invention. Upper diagram in Figure 3 shows how the congestion of the vehicle 5 at time t _1, the diagram below in Figure 3, the state of congestion of the vehicle 5 from time t ₁ at time t ₂ after a predetermined time Shown.

In the following description, when it is simply called the inflow path 1, it means the inflow path 1 on the south side of the intersection.

Referring to the upper side of FIG. In FIG. 3, the vehicle 5 mounted with the vehicle-mounted device (probe car 5a: solid vehicle) is, at time t _1, stuck in traffic in the inflow path 1, the most upstream congestion Suppose you stop at a position. In this case, it is assumed that the position when the probe wheel 5a is stopped is the position of 650 m from the stop line 4. In this case, the end position of the traffic jam is the position from the stop line 4 to 650 m, and the traffic jam length (distance from the stop line 4 to the end position of the traffic jam) of the vehicle 5 in the inflow path 1 is 650 m.

The end position of the traffic jam will be described as the position of the vehicle 5 stopped on the most upstream side of the traffic jam, but the end position of the traffic jam is a predetermined speed (for example, 10 km / h) on the most upstream side of the traffic jam. It may be the position of the vehicle 5 traveling below.

Referring to the lower diagram in FIG. 3, the probe car 5a is passed through the intersection (straight, right turn, or left) to exit the traffic jam, the time t _2, the transmitting the probe information to the road device 40 And. The probe information transmitted from the probe vehicle 5a includes time information, position information of the vehicle 5 at that time, and the like. Accordingly, the control unit 11 of the control device 10, if the time t ₂ later, based on the acquired probe information, the actual (650 meters from the stop line 4) tail position of the congestion at time t ₁ to recognize accurately Can be done.

However, at the point of time t _2, the information of the end position of the congestion at the time t ₁ is already past of information, the end position of the congestion is, at the point of time t _2, in many cases you've already changed .. In the lower diagram in FIG. 3, at the point of time t _2, the situation is shown when the tail position of the congestion is has changed from the stop line 4 to the position of 750 meters.

Here, the longer the congestion length, time to exit the jam from being caught up in the probe cars 5a traffic congestion is increased (i.e., the time from time t ₁ to time t ₂ longer). Therefore, the longer the traffic jam length, the older the information at the end of the traffic jam in the probe information.

On the other hand, as shown in the upper part of FIG. In FIG. 3, at time _{t 1,} the three sensors 32, sensing information (traffic volume at each point (300 meters from the stop line 4, 500 meters, 1000 m), occupancy rate Information) has been acquired. At time t _1, between the the sensing information, the tail position of the congestion (650 meters from the stop line 4), there should be a correlation.

By using this correlation, it is considered that the end position of the traffic jam at the current time can be accurately estimated from the sensed information at the current time. The above is the basic idea of the present invention.

<Operation explanation>
Next, the processing of the control unit 11 in the control device 10 will be described. The processing of the control unit 11 is roughly classified into two types of processing. The first type is a process of learning (determining) the correlation between the end position of the traffic jam and the sensed information in the past time, and the second type is the process of learning (determining) the end position of the traffic jam based on these correlations at the current time. It is a process to estimate.

FIG. 4 is a diagram showing a process for learning (determining) the correlation between the end position of the traffic jam and the sensed information at the past time. FIG. 5 is a diagram showing processing when estimating the end position of the traffic jam based on the correlation at the current time.

First, with reference to FIG. 4, a process for learning (determining) the correlation between the end position of the traffic jam and the sensed information at the past time will be described.

With reference to FIG. 4, first, the control unit 11 acquires and collects the probe information transmitted from the vehicle 5 and stores it in the storage unit 12 (step 101). Further, the control unit 11 acquires and collects the detection information acquired by the three sensing devices 30 in parallel with the collection of the probe information, and associates the sensing information with the time when the sensing information is acquired. It is stored in the storage unit 12 (step 102).

In the description here, the case where the sensing information acquired by the three sensing devices 30 is used for the processing of the control unit 11 will be described, but one, two, or four or more sensing devices 30 will be used. The acquired sensing information may be used for the processing of the control unit 11 (the same applies in FIG. 5). As the number of sensing devices 30 for acquiring sensing information increases, the accuracy of correlation determination (in FIG. 5, the accuracy of estimating the end position of the traffic jam) can be improved.

Further, the sensing information used for the processing of the control unit 11 may be one of the traffic volume information of the vehicle 5, the occupancy rate information of the vehicle 5, and the speed information of the vehicle 5, and may be two or more. (Similar in FIG. 5). As the types of information used for processing increase, the accuracy of correlation determination (in FIG. 5, the accuracy of estimating the end position of traffic congestion) can be improved.

After collecting the probe information and the sensing information, the control unit 11 next determines whether or not a predetermined time has elapsed since the previous process such as correlation learning was executed (step 103). This predetermined time indicates a cycle in which processing such as learning the correlation is executed, and this cycle is, for example, several hours or several days.

When the predetermined time has elapsed from the previous time (YES in step 103), the control unit 11 determines the end position of the traffic jam at the past time based on each collected probe information (step 104).

For example, as shown in the upper diagram of FIG 3, at the past time t _1, the probe car 5a is stuck in traffic in the inflow path 1, and stops at the most upstream position of the traffic congestion (or below a predetermined speed It became). It is assumed that the probe information has already been acquired (see the lower side of FIG. 3). In this case, the probe car 5a is stopped at the position of the most upstream congestion (or a predetermined speed falls below) a position (650 meters from the stop line 4), it is determined as the last position of the traffic congestion in the past time t ₁ To.

When the probe vehicle 5a is involved in a traffic jam, there is always a moment when the probe vehicle 5a is at the end of the traffic jam. Therefore, the control unit 11 uses one probe information to determine the end position of the traffic jam at a past time. One can be determined. It should be noted that the end position of the traffic jam at the past time determined based on the probe information has high accuracy and can be regarded as a true value without any problem.

After determining the end position of the traffic jam, the control unit 11 next extracts the end position of the traffic jam at the past time and the sensed information at the same time as the past time as teacher data (step 105). For example, the control unit 11, the end position of the traffic congestion in the past time t ₁ and (650 meters from the stop line 4), (also past time other) for extracting the sense data at past time t ₁ as teacher data ..

The end position and detection information of the traffic jam extracted as teacher data are typically the end position and detection information of all traffic jams, but the end position and detection information of some traffic jams are used as teacher data. It may be extracted.

Next, the control unit 11 associates the end position of the traffic jam extracted as the teacher data with the sensed information, stores it as an actually measured value, and creates a database (step 106). Next, the control unit 11 learns the correlation between the end position of the traffic jam at the past time and the sensed information at the past time and the same time by a learning function such as a neural network (step 107).

Next, the control unit 11 generates an estimation parameter based on the learned correlation (step 108). This estimation parameter is a parameter used in the process of estimating the end position of the traffic jam at the current time from the sensing information at the current time, and is, for example, a weight value. By using such an estimation parameter, the end position of the traffic jam at the current time can be appropriately estimated.

Next, the control unit 11 applies the generated estimation parameter to the sensing information at the past time in order to evaluate whether the generated estimation parameter is accurate, and the traffic jam at the same time as the past time. Estimate the end position of (step 109).

In step 109, the control unit 11 may estimate the end position of the traffic jam from all the past sensed information that can estimate the end position of the traffic jam, or some samples of all the past sensed information. May be extracted to estimate the end position of the traffic jam.

After estimating the end position of the traffic jam, the control unit 11 then reads the end position (true value) of the actual traffic jam at the past time from the database, and the end position (true value) of the actual traffic jam and the estimation parameter. Calculate the difference from the end position of the traffic jam estimated by (step 110).

Next, the control unit 11 determines whether or not the average value of the calculated differences is equal to or less than a predetermined threshold value (that is, evaluates the quality of the estimated parameter) (step 111). When the mean value of the difference is equal to or less than a predetermined threshold value (YES in step 111), that is, when it is determined that the quality of the estimated parameter is equal to or higher than a certain standard, the control unit 11 newly sets the current estimated parameter. Overwrite (update) with various estimation parameters (step 112). This allows the estimated parameters to be updated appropriately.

On the other hand, when the average value of the differences exceeds a predetermined threshold value (NO in step 111), that is, when it is determined that the quality of the estimated parameters does not meet a certain criterion, the control unit 11 overwrites the estimated parameters (NO). The process ends without performing update).

Next, with reference to FIG. 5, a process for estimating the end position of the traffic jam based on the correlation at the current time will be described.

First, the control unit 11 acquires and collects the sensing information acquired by the three sensing devices 30, and stores the sensing information and the time when the sensing information is acquired in the storage unit 12 (step 201). ).

Next, the control unit 11 determines whether or not a predetermined time has elapsed since the end position of the traffic jam at the current time was estimated in the previous time (step 202). This predetermined time indicates a cycle in which the end position of the traffic jam at the current time is estimated, and this cycle is, for example, several tens of seconds or several minutes.

When the predetermined time has elapsed from the previous time (YES in step 202), the control unit 11 reads the sense information at the current time from the storage unit 12 (step 203).

Next, the control unit 11 estimates the end position of the traffic jam at the current time from the sensed information at the current time by applying an estimation parameter to the sensed information at the current time (step 204). That is, the control unit 11 estimates the end position of the traffic jam at the current time from the sensing information at the current time based on the correlation. Since the control unit 11 can estimate the end position of the traffic jam at the current time, it can also estimate the length of the traffic jam at the current time.

The control unit 11 controls the display time of the signal light 20 (blue, yellow, red, arrow signal light) based on the information of the end position of the traffic jam at the estimated current time. Further, the control unit 11 has congestion information for transmission to the probe vehicle 5a based on the information on the end position of the congestion at the estimated current time (information on the length of the congestion, information on the time required to clear the congestion, and congestion. Route information, etc. to avoid The traffic congestion information generated by the control device 10 is transmitted from the control device 10 to the road device 40, and then transmitted from the road device 40 to the probe vehicle 5a.

<Action, etc.>
As described above, in the control device 10 according to the present embodiment, there is a correlation between the end position (true value) of the traffic jam at the past time acquired from the probe information and the sensed information at the same time as the past time. Learned (judged). Then, based on this correlation, the end position of the traffic jam at the current time is estimated from the sensed information at the current time. By estimating the end position of the traffic jam by such a method, the accuracy of estimating the end position of the traffic jam can be improved.

In particular, in the present embodiment, even when the distance between the sensing devices 30 is long, the end position of the traffic jam between the sensing devices 30 can be accurately estimated at the current time. Further, in the present embodiment, even if the traffic jam exceeds the position of the sensing device 30 (1000 m from the stop line 4) installed at the uppermost stream of the inflow path 1, the traffic jam is accurately caused at the current time. The end position can be estimated. Further, in the present embodiment, the end position of the traffic jam can be appropriately estimated even in a situation where the penetration rate of the probe vehicle 5a is low (because the probe information may be collected as past information).

Further, in the present embodiment, since the end position of the traffic jam can be accurately estimated, the display time of the signal light 20 can be set to an appropriate length according to the length of the traffic jam. As a result, traffic congestion can be alleviated. Further, in the present embodiment, since the end position of the traffic jam can be accurately estimated, accurate traffic jam information can be generated and transmitted to the probe vehicle 5a. As a result, the quality of the traffic jam information providing service can be improved.

1 ... Inflow path 2 ... Outflow path 5 ... Vehicle 10 ... Control device 11 ... Control unit 20 ... Signal light 30 ... Sensing device 40 ... Road device

Claims (6)

Get the driving information sent from the vehicle and
Acquire the sensing information obtained by sensing the vehicle,
Based on the driving information, the end position of the traffic jam at the past time is determined.
The correlation between the end position of the traffic jam at the past time and the sensed information at the same time as the past time is determined.
A control unit that generates an estimation parameter based on the correlation , applies the estimation parameter to the sensing information at the current time, and estimates the end position of the traffic jam at the current time .
The control unit evaluates the quality of the estimated parameter by applying the estimated parameter to the sensed information at the past time and estimating the end position of the traffic jam at the same time as the past time. , A control device that updates the estimated parameters according to the evaluation result . The control device according to claim 1.
The control unit is a control device that learns the correlation by a learning function. The control device according to claim 1 or 2 .
The control unit is a control device that controls the display time of the signal light based on the end position of the traffic jam at the estimated current time. The control device according to any one of claims 1 to 3 .
The control unit is a control device that generates traffic jam information for transmission to the vehicle based on the estimated end position of the traffic jam at the current time. Get the driving information sent from the vehicle and
Acquire the sensing information obtained by sensing the vehicle,
Based on the driving information, the end position of the traffic jam at the past time is determined.
The correlation between the end position of the traffic jam at the past time and the sensed information at the same time as the past time is determined.
Based on the correlation, an estimation parameter is generated, and the estimation parameter is applied to the sensing information at the current time to estimate the end position of the congestion at the current time.
It ’s an estimation method.
In the step of estimating the end position of the traffic jam at the current time,
The quality of the estimated parameter is evaluated by applying the estimated parameter to the sensed information at the past time and estimating the end position of the traffic jam at the same time as the past time, and according to the evaluation result. An estimation method for updating the estimation parameters . Steps to acquire driving information transmitted from the vehicle and
The step of acquiring the sensing information obtained by sensing the vehicle, and
Based on the driving information, the step of determining the end position of the traffic jam at the past time and
A step of determining the correlation between the end position of the traffic jam at the past time and the sensed information at the same time as the past time.
With the step of generating an estimation parameter based on the correlation and applying the estimation parameter to the sensing information at the current time to estimate the end position of the traffic jam at the current time.
Is a program that causes a computer to execute
In the step of estimating the end position of the traffic jam,
The quality of the estimated parameter is evaluated by applying the estimated parameter at the current time to the sensing information at the past time and estimating the end position of the congestion at the same time as the past time. A program that causes a computer to perform steps to update the estimated parameters according to the evaluation result .

Images