JP5037237B2 - Traffic signal control device and outflow traffic flow prediction method - Google Patents

Description

  The present invention relates to an autonomous distributed traffic signal control device and an outflow traffic flow prediction method in an autonomous distributed traffic signal control device.

The autonomous decentralized traffic signal control device controls traffic signals at its own intersection. For example, it predicts traffic demand at its own intersection in the near future such as several minutes ahead, and responds to the predicted traffic demand in detail. Perform optimal signal control. Specifically, by exchanging predicted outflow traffic data with other intersection control devices, it predicts time-series changes in arrival traffic flow at its own intersection, and optimizes signal control at its own intersection. The parameter is determined and signal control is performed (for example, refer to Patent Document 1).
JP 2005-182219 A

  However, the conventional prediction of outflow traffic flow as shown in Patent Document 1 does not reflect the influence of cross traffic. In actual traffic, for example, when turning right, the vehicle is blocked by the oncoming straight vehicle and cannot smoothly turn right, so the number of outflows in the right turn direction cannot be handled in the same way as the number of outflows in the straight direction. More specifically, when there are many oncoming straight cars, the progress is completely blocked and the number of outflows becomes zero, and when there are not so many, there is no oncoming car because it turns right through the gap of the oncoming straight car Compared to the number of outflows. Similarly, when turning left, since the pedestrian crossing the pedestrian crossing provided in the inflow path in the left turn direction is blocked, the number of outflows in the left turn direction is handled in the same way as the outflow number in the straight direction. Can not.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to enable highly accurate outflow traffic flow prediction by considering the traffic volume of other traffic that crosses in the outflow direction.

The first invention for solving the above-described problems is
Communication means (for example, the communication control unit 200 in FIG. 10) that mutually transmits and receives outflow traffic flow information to and from other traffic signal control devices at other intersections, and arrival traffic flow prediction that predicts arrival traffic flow at the own intersection Means (for example, arrival traffic flow prediction unit 110 in FIG. 10) and the outflow traffic flow for predicting the outflow traffic flow from the own intersection by predicting the number of outflows by the outflow direction from each inflow path of the own intersection to the other road An autonomous decentralized traffic signal control device (for example, FIG. 10) that includes a prediction means (for example, the outflow traffic flow prediction unit 130 in FIG. 10) and controls the signal at the own intersection by changing the signal control parameter of the own intersection. A traffic signal control device 20),
The outflow traffic flow predicting means, when predicting the number of outflows according to the outflow direction from each inflow path of the intersection to the other road, outflow of the object to be predicted among the crossing traffic crossing in each outflow direction determined in advance. It is a traffic signal control device having outflow number correction means for correcting the number of outflows in the outflow direction based on the traffic volume of the crossing traffic corresponding to the direction.

In addition, the seventh invention,
Information on the outflow traffic flow is sent to and received from other traffic signal control devices at other intersections, and the arrival traffic flow to the own intersection and the outflow traffic flow from the own intersection are predicted as needed. A method for predicting outflow traffic flow in an autonomous decentralized traffic signal control device that performs signal control of its own intersection by changing signal control parameters,
When predicting the number of outflows by direction of outflow from each inflow road to the other road at the intersection, the traffic of the crossing traffic corresponding to the outflow direction of the prediction target among the crossing traffic crossing in each predetermined outflow direction This is an outflow traffic flow prediction method in which the number of outflows by outflow direction is predicted by correcting the number of outflows in the outflow direction based on the volume.

  According to the first or seventh invention, in the autonomous distributed traffic signal control, when predicting the number of outflows by the outflow direction from each inflow path of the intersection to the other road, Based on the traffic volume of cross traffic, the number of outflows in the outflow direction is corrected. When there is other traffic that intersects the outflow direction of the vehicle, the progress is blocked by this crossing traffic, and the number of outflows in the outflow direction decreases or becomes zero. For example, at the time of a right turn, the progress is blocked by the oncoming straight vehicle, and at the time of a left turn, the progress is blocked by a pedestrian crossing the inflow path in the left turn direction. However, as in the present invention, it is possible to predict the outflow traffic flow with high accuracy by correcting and predicting the number of outflows according to the traffic volume of the cross traffic.

The second invention is the traffic signal control device of the first invention,
The outflow traffic flow predicting means predicts an outflow number temporary prediction means (for example, an outflow traffic flow prediction unit 130 in FIG. )
The spill number correction means predicts the tentative spill number predicted by the spill number provisional prediction means in the prediction target spill direction by the spill number tentative prediction means of the intersection traffic corresponding to the prediction target spill direction. Correct based on the number of spills,
It is a traffic signal control device.

  According to the second aspect of the invention, the number of outflows for each outflow direction from each inflow path at the intersection to the other road is predicted as a temporary outflow number. Then, the number of provisional outflows in the outflow direction to be predicted is corrected based on the number of provisional outflows of the corresponding intersection traffic.

The third invention is the traffic signal control device of the first or second invention,
When the outflow direction to be predicted is a right turn direction, the number of outflow correction means is based on the traffic volume of an oncoming straight vehicle and / or an oncoming left turn that is predetermined as a crossing traffic with respect to the right turn direction. It is a traffic signal control apparatus which has the 1st correction | amendment means (For example, the outflow traffic flow estimation part 130 of FIG. 10) which correct | amends the number of outflows to.

  According to the third aspect of the present invention, when the outflow direction to be predicted is the right turn direction, based on the traffic flow of the oncoming straight vehicle and / or the oncoming left turn vehicle that is predetermined as the crossing traffic for the right turn direction, The number of outflows in the right turn direction is corrected. Therefore, as a typical example in which the progress is blocked by the crossing traffic, it is possible to predict the outflow traffic flow with higher accuracy in consideration of the oncoming straight vehicle when turning right.

A fourth invention is the traffic signal control device according to any one of the first to third inventions,
The outflow number correction means reduces or suppresses the outflow number in the outflow direction of the prediction target based on the traffic volume for the past predetermined time from the prediction target time of the intersection traffic corresponding to the outflow direction of the prediction target. It is.

  According to the fourth aspect of the invention, the number of outflows in the outflow direction of the prediction target is reduced or suppressed based on the traffic volume for the past predetermined time from the prediction target time of the intersection traffic corresponding to the outflow direction of the prediction target. .

A fifth invention is the traffic signal control device according to any one of the first to fourth inventions,
The outflow number correction means corrects the outflow number in the outflow direction when the traffic volume for a predetermined time in the past from the prediction target time of the intersection traffic corresponding to the outflow direction of the prediction target satisfies a predetermined low traffic amount condition. It is a traffic signal control device which corrects when it does not satisfy.

  According to the fifth aspect of the present invention, when the traffic volume for the past predetermined time from the prediction target time of the intersection traffic corresponding to the prediction target outflow direction satisfies the predetermined low traffic volume, the outflow in the outflow direction It is corrected when the number is not corrected and does not satisfy. For example, in the case of a right turn, even if there are oncoming straight vehicles, if the number is small, the oncoming vehicles will not be blocked from traveling in the right turn direction. For this reason, when the traffic flow of the crossing traffic is so small as not to prevent the prediction target from flowing in the outflow direction, the number of outflows in the outflow direction need not be corrected.

A sixth invention is the traffic signal control device according to any one of the first to fifth inventions,
Pedestrian crossing pedestrian prediction means for predicting the traffic volume of the pedestrian crossing pedestrian at the intersection (for example, the outflow traffic flow prediction unit 130 of FIG. 10),
When the outflow direction to be predicted is a right turn direction or a left turn direction, the number of outflow correction means is a crosswalk pedestrian on the outflow path in the outflow direction that is predetermined as the crossing traffic for the right turn direction or the left turn direction. Having a second correcting means for correcting the number of outflows in the outflow direction based on the traffic volume;
It is a traffic signal control device.

  According to the sixth aspect of the present invention, when the outflow direction to be predicted is the right turn direction or the left turn direction, the pedestrian crossing pedestrian on the outflow path in the outflow direction predetermined as the intersection traffic for the right turn direction or the left turn direction. The number of outflows in the outflow direction is corrected based on the traffic volume. Therefore, as another example in which the progress is blocked by cross traffic, it is possible to predict the outflow traffic flow with higher accuracy in consideration of crossing pedestrians.

  According to the present invention, in the autonomous decentralized traffic signal control, when predicting the number of outflows according to the outflow direction from each inflow path to the other road at the intersection, the traffic volume of the intersection traffic in the outflow direction to be predicted is calculated. Based on this, the number of outflows in the outflow direction is corrected. As a result, it is possible to predict the outflow traffic flow with high accuracy.

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

[System configuration]
FIG. 1 is an overall configuration diagram of an autonomous distributed traffic signal control system 1 of the present embodiment. The autonomous decentralized traffic signal control system 1 is configured by connecting a central management device 10 installed in a control center and a plurality of traffic signal control devices 20 installed at each intersection via a transmission line N. . The traffic signal control device 20 is connected to a plurality of traffic signals 30 and a vehicle detector 40 provided at an intersection where the device is installed (own intersection). The traffic signal control device 20 transmits / receives data to / from other traffic signal control devices 20 such as adjacent intersections, and each traffic signal device 30 provided at its own intersection based on a vehicle detection signal from the vehicle detector 40. To control.

[intersection]
FIG. 2 is a layout diagram of intersections to be controlled by the traffic signal control device 20. This intersection is a cross intersection having four inflow paths A to D, and four traffic signals 30 and a vehicle detector 40 are installed in each of the inflow paths A to D. The figure is an example of an intersection, and is not limited to this intersection, and can be applied to intersections of other shapes such as a three-way intersection or a T-shaped intersection.

  The signal display at the intersection is a two-display system. FIG. 3 shows an example of the present. In the figure, a solid arrow indicates a flow line of a vehicle to which a right of traffic is given, and a broken arrow indicates a flow line of a pedestrian to which the right of right is given. That is, at this intersection, the present indication φ1 that gives the right of traffic to the inflow paths B and D and the present indication φ2 that gives the right of passage to the vehicle traffic and the pedestrian traffic of the inflow paths A and C are alternately displayed. Is done. This figure is an example of the presenting method, and the present invention is not limited to the two presenting methods, and a multiple presenting method having three or more presenting methods may be used.

[Outline of signal control]
FIG. 4 is a schematic diagram of signal control in the traffic signal control device 20. The traffic signal control device 20 first predicts the traffic flow arriving at its own intersection based on (1) the predicted outflow traffic flow information received from the traffic signal control device 20 at the adjacent intersection. Next, (2) based on the predicted arrival traffic flow (predicted arrival traffic flow), control parameters (cycle length C, split and offset) of each traffic signal 30 at the own intersection are calculated. Then, (3) according to the calculated control parameter, each traffic signal 30 at the own intersection is controlled.

  Further, based on the predicted arrival traffic flow (predicted arrival traffic flow) and the calculated control parameter (signal control parameter) of each traffic signal 30, the traffic flow flowing out from the own intersection (outflow traffic flow) is predicted. Then, the traffic flow information predicted to flow out (predicted outflow traffic flow information) is transmitted to the traffic signal control device 20 at each adjacent intersection.

[principle]
(1) Runoff prediction The outflow traffic flow prediction principle will be explained. 5 and 6 are diagrams showing the principle of prediction of outflow traffic flow, and show the inflow path A or C in FIG. As shown in FIG. 1, the inflow paths A and C have a two-lane straight turn left lane and a one-lane right turn lane. Further, the saturated traffic flow is “0.5 [unit / s]” equally in any lane.

  A vehicle that has arrived at the inflow path travels straight, turns left or right in the intersection, and flows out from the inflow path (direction) corresponding to the travel direction. In which direction the vehicle that has arrived on the inflow path travels is determined stochastically. Specifically, the probability of going straight ahead (straight forward rate) is “0.7 (70%)”, the probability of going left turn (left turn rate) is “0.1 (10%)”, and turns right The probability of proceeding (right turn rate) is “0.2 (20%)”.

(1-1) When there is no right of traffic FIG. 5 is an explanatory diagram of the outflow prediction when there is no right of traffic in the inflow channel. In the figure, in order from the left side in the figure, the arrival traffic flow of the inflow path, the planned display, and the number of staying in each lane are shown as the time axis t in the lower direction in the figure. When there is no right to pass, the vehicle that has arrived at the inflow path stays in the inflow path as it is. That is, all the vehicles that have arrived at the inflow path at time t when the scheduled display is “red” stay in the inflow path. More specifically, of the number of arrivals, the number of vehicles having a ratio of “0.8 (= straight ahead rate“ 0.7 ”+ left turn rate“ 0.1 ”)” arrives at the straight turn left lane, and “0.2 (= The right turn rate)) will arrive in the right turn lane. The number of staying lanes at time t is the number obtained by adding the number of arrivals to the number of staying lanes at the immediately preceding time t.

For example, the number of arrivals at time t ₁ when the scheduled display is “red” is “0.5 [unit]”. That is, “0.4 (= 0.5 × 0.8) [unit]” arrives at the straight turn left lane, and “0.1 (= 0.5 × 0.2) [unit]” enters the right turn lane. arrive. Then, the number of staying in each lane at time t ₁ is obtained by adding the number of arrivals “0.4 [units]” to the number of staying “3.2 [units]” at the immediately preceding time t ₀ for the straight turn left lane. 3.6 [units] ”, and for the right turn lane,“ 0.3 [units] is obtained by adding the arrival number “0.1 [units]” to the number of remaining units “0.2 [units]” at the previous time t ₀ . ] ”.

(1-2) When there is a right of traffic FIG. 6 is an explanatory diagram of outflow prediction when there is a right of traffic in the inflow channel. In the figure, in order from the left side in the figure, the arrival traffic flow of the inflow path, the planned display, the number of staying in each lane, and the outflow traffic flow are shown as time t in the downward direction in the figure. When there is a right of passage, the vehicle that has arrived at the inflow path travels along the intersection together with the vehicle that has stayed in the inflow path, and flows out from the other inflow path. That is, at time t when the scheduled display is “blue”, the vehicle that has arrived on the inflow route is “0.8 (= straight-ahead rate“ 0.7 ”), similarly to the case where there is no right of traffic shown in FIG. + Left turn rate “0.1”) ”arrives at the straight turn left lane, and“ 0.2 (right turn rate) ”arrives at the right turn lane. Here, the vehicle arriving in each lane is considered to be temporarily staying in the lane together with the vehicle staying in the lane, and the number obtained by adding this arrival number to the staying number at the immediately preceding time t is This is called “temporary residence”.

  Then, the vehicle temporarily staying in each lane travels in a straight direction, a left turn or a right turn direction, and flows out from another inflow path. That is, from the straight ahead left turn lane, the temporarily staying vehicle flows out in the straight direction or the left turn direction. Specifically, from the straight turn left turn lane, the number of vehicles staying at a ratio of “0.7 (straight forward rate) /0.8 (straight forward rate“ 0.7 ”+ left turn rate“ 0.1 ”)” is straight ahead. The number of vehicles with a ratio of “0.1 (left turn rate) /0.8 (straight ahead rate“ 0.7 ”+ left turn rate“ 0.1 ”)” flows out in the left turn direction. Moreover, all the vehicles that have stayed temporarily flow out from the right turn lane in the right turn direction.

  However, the number of outflows from each lane is determined so as not to exceed the saturation traffic flow of the lane. In other words, if the number of temporarily staying is less than or equal to the saturated traffic flow, the number of temporarily staying is the number of outflows from the lane, and if the number of temporarily staying exceeds the saturation traffic flow, the number of saturated traffic flows from the lane. The number of outflows. In the examples of FIGS. 1, 5, and 6, the straight turn left lane has two lanes, so the lane number “2” is added to the saturated traffic flow “0.5 [unit / s]” per lane. Multiplyed by “1.0 [unit / s]”.

For example, the arrival number at time t ₃ when the planned actual time is “blue” is “0.5 [unit]”. That is, among the arriving vehicles, “0.4 (= 0.5 × (straight ahead rate“ 0.7 ”+ left turn rate“ 0.1 ”))” [car] ”arrives in the straight left turn lane, and“ 0 .1 (= 0.5 × Right turn rate “0.2”) [unit] ”arrives on the right turn lane. Then, in a straight left turn lane, the sum of the arrival number "0.4 [units]" to the residence number at time t ₂ of the previous "3.6 [units],""4.0[units]" is a temporary residence number However, since the temporarily staying number “4.0 [units]” exceeds the saturated traffic flow “1.0 [units / s] (= 0.5 × 2 [lanes])”, it flows out from the straight turn left lane. The number is “1.0 [unit]”. Then, out of the number of outflows “1.0 [unit]”, “0.9 (≈1.0 × 0.7 / 0.8) [unit]” proceeds in the straight direction and flows out. .1 (≈1.0 × 0.1 / 0.8) [unit] ”proceeds in the left turn direction and flows out. Therefore, the residence number at time _{t 3} in the straight left turn lane is "3.0 (= 4.0-1.0) [units]."

On the other hand, in the right turn lane, “0.4 [unit]”, which is obtained by adding the arrival number “0.1 [unit]” to the staying number “0.3 [unit]” at the time t ₂ immediately before, becomes the temporary residence number. However, since the temporary residence number “0.4 [unit]” does not reach the saturated traffic flow “0.5 [unit / s]”, the number of outflows from the right turn lane is “0.4 [unit]”. Become. And all of this outflow number “0.4 [unit]” flows out in the right turn direction. Therefore, the residence number at time _{t 3} in the right turn lane is "0.0 [units]."

  However, in consideration of the time Δt required for the vehicle to pass through the intersection, the vehicle that has flowed out from each lane of the inflow path at time t flows out of the traveling direction at time (t + Δt) after the crossing time Δt. I will do it.

(2) Correcting the number of outflows according to crossing traffic Further, in the present embodiment, traffic (crossing traffic) that crosses out the number of outflows by traveling direction from each inflow path calculated as described above in the traveling direction. Correct according to traffic volume. Specifically, the number of outflows in the direction of travel is corrected according to (1) an oncoming straight vehicle at the time of a right turn and (2) a pedestrian crossing pedestrian at the time of a right or left turn.

(2-1) Opposite Straight Car at Right Turn FIG. 7 is a diagram for explaining the correction of the number of outflows according to the oncoming straight car at the right turn. In the intersection shown in the figure, when traveling in a right turn direction from a certain inflow path, the opposite straight vehicle and the opposite left turn vehicle become traffic (cross traffic) crossing in the traveling direction (right turn direction). And according to the traffic volume of this crossing traffic, the number of outflows to the said advancing direction (right turn direction) is correct | amended.

  Specifically, when the traffic volume of the crossing traffic in a past predetermined time (for example, 3 seconds) from the prediction target time t exceeds a predetermined threshold (for example, 5 vehicles), the number of outflows at the time t is set to “0”. " The traffic volume of the crossing traffic is the total number of staying vehicles and arrivals in the lane corresponding to the crossing traffic (that is, temporarily staying). For example, when traveling in the right turn direction from the inflow path C, the crossing traffic for the traveling direction (right turn direction) is a vehicle that travels straight from the inflow path A or in the left turn direction (straight-turn left turn vehicle). The number of outflows in the right turn direction from the inflow path C at the predicted time t is the total number of stays and arrivals in the straight left turn lane of the inflow path A between the predicted time t and a predetermined past time. If it exceeds the threshold number, it is corrected to “0”.

(2-2) Pedestrian crossing pedestrian at the time of right / left turn FIG. 8 is a diagram for explaining the correction of the number of outflows according to the pedestrian (pedestrian) of the pedestrian crossing at the time of right / left turn. In the intersection shown in the figure, when traveling in a right turn or left turn direction from a certain inflow path, a crossing pedestrian on the inflow path in the travel direction (right turn or left turn direction) becomes cross traffic with respect to the travel direction. And according to the traffic volume of this crossing traffic, the number of outflows in the traveling direction is corrected. Specifically, during the time corresponding to the number of crossing pedestrians crossing in the traveling direction, the number of outflows in the traveling direction is set to “0”. At this time, the number of crossing pedestrians is determined so as to be proportional to the number of people staying in the outflow source inflow path at the timing when the current display is switched. Then, the number of outflows in the traveling direction is set to “0” until the time tu corresponding to the determined crossing pedestrian elapses from the timing when the current display is switched.

  FIG. 9 shows an example of the relationship between the number of crossing pedestrians and the crossing time tu. In the figure, a graph is shown in which the horizontal axis represents the number of crossing pedestrians and the vertical axis represents the crossing time tu. According to the figure, it is determined that the crossing time tu becomes longer as the number of crossing pedestrians increases. Further, an upper limit value tumax and a lower limit value tumin are determined for the crossing time tu. The upper limit value tumax is determined to be shorter than, for example, the currently displayed step number of seconds S that gives the right of passage to the outflow source inflow path.

  For example, in FIG. 8, when making a left turn from the inflow path C, a crossing pedestrian in the inflow path D becomes a crossing traffic with respect to the traveling direction (left turn direction). In this case, the number of crossing pedestrians on the inflow path D, which is a crossing traffic, is determined so as to be proportional to the number of staying in the inflow path C at the timing when the current display is switched. Then, the number of outflows in the left turn direction from the inflow path C is corrected to “0” for a time tu corresponding to the number of crossing pedestrians in the inflow path D from the timing when the current display is switched.

  Note that the number of crossing pedestrians, which are cross traffic, is determined based on the number of staying in the outflow source inflow path (for example, inflow path C in FIG. 8), but staying in the lane corresponding to the target traveling direction. It may be based on the number. Specifically, in FIG. 8, when traveling in the left turn direction from the inflow path C, the number of crossing pedestrians in the inflow path D is determined based on the number of staying in the straight left turn lane of the inflow path C. Alternatively, it may be based on the measured staying number calculated based on the sensing result of the sensing signal by the vehicle detector 40 provided in the corresponding inflow path. Furthermore, a pedestrian detector (sensor) that counts pedestrians may be installed in front of the pedestrian crossing, and the number of crossing pedestrians measured based on the detection result by the pedestrian detector may be used. .

[Configuration of traffic signal controller]
FIG. 10 is a block diagram illustrating an internal configuration of the traffic signal control device 20. According to the figure, the traffic signal control device 20 includes a processing unit 100, a communication control unit 200, and a storage unit 300.

  The processing unit 100 is a program or data stored in the storage unit 300, or data received from an external device (mainly, traffic signal control device 20 at an adjacent intersection) via the communication control unit 200 (predicted outflow traffic flow information). ) Etc., various processes such as overall control of the traffic signal control device 20 are performed. The processing unit 100 is realized by a CPU, for example. The processing unit 100 includes an arrival traffic flow prediction unit 110, a control parameter calculation unit 120, an outflow traffic flow prediction unit 130, and a signal control unit 140.

  Note that the configuration of an intersection (self-intersection) to be controlled by the traffic signal control device 20 is defined by the self-intersection configuration table 321. FIG. 11 shows an example of the data configuration of the own intersection configuration table 321. According to the figure, the self-intersection configuration table 321 stores, for each inflow path 321a at the self-intersection, the lane 321b and the number of lanes 321c constituting the inflow path and the pedestrian crossing 321d are associated with each other.

  The arrival traffic flow prediction unit 110 calculates a traffic flow predicted to arrive (predicted traffic flow) based on the predicted outflow traffic flow data 331 at the adjacent intersection received from the traffic signal control device 20 at the adjacent intersection. Specifically, based on the predicted outflow traffic flow data 331 of the adjacent intersection, the outflow traffic flow from the inflow path connected to the inflow path of the own intersection among the inflow paths of the adjacent intersection from the adjacent intersection to the own intersection The travel time will be delayed and the traffic will arrive at the intersection. The arrival traffic flow prediction unit 110 repeatedly executes the calculation of the predicted arrival traffic flow at predetermined time intervals.

  The predicted outflow traffic flow data 331 is data of a traffic flow (predicted outflow traffic flow) predicted to flow out from the intersection. FIG. 12 shows an example of the predicted outflow traffic flow data 331. According to FIG. 6A, the predicted outflow traffic flow data 331 includes predicted outflow traffic flow data 331A at its own intersection and predicted outflow traffic flow data 331B at each adjacent intersection. The predicted outflow traffic flow data 331A of the own intersection is generated by the outflow traffic flow prediction unit 130, and the predicted outflow traffic flow data 331B of the adjacent intersection is data received from the other traffic signal control device 20.

According to FIG. 5B, the predicted outflow traffic flow data 331 stores the number of outflows 331b from each inflow path of the corresponding intersection in association with each time 331a within the prediction target time range. Time 331a after the current time t ₀ is a time range prediction target predetermined time (e.g., 200 seconds) of between times t _n of the predetermined time intervals (e.g., intervals of one second) is the successive times in .

The predicted arrival traffic flow data 332 is data of traffic flow (arrival traffic flow) predicted to arrive at the own intersection. FIG. 13 shows an example of the data structure of the predicted arrival traffic flow data 332. According to the figure, the predicted arrival traffic flow data 332 stores the number of arrivals 332b to each inflow path at the own intersection in association with each time 332a within the time range to be predicted. Time 332a, similar to the predicted outflow traffic flow data 331, from the current time _{t 0} is a time range prediction target until time _{t n} after a predetermined time (e.g., after 200 seconds), a predetermined time interval (e.g., 1 (Second interval).

  Based on the predicted arrival traffic flow data 332 generated by the arrival traffic flow prediction unit 110, the control parameter calculation unit 120 calculates the control parameters (cycle length C, split and offset) of each traffic signal 30 at the own intersection. The control parameter calculation unit 120 repeatedly executes this signal control parameter calculation at predetermined time intervals.

  The calculated control parameter is stored as signal control parameter data 341. FIG. 14 shows an example of the signal control parameter data 341. According to the figure, the signal control parameter data 341 stores a cycle length 341a, a split 341b, and an offset 341c, which are signal control parameters. The signal control parameter data 341 is updated every time the control parameter calculation unit 120 calculates the control parameter.

  Based on the predicted arrival traffic flow calculated by the arrival traffic flow prediction unit 110 and the signal control parameter calculated by the control parameter calculation unit 120, the outflow traffic flow prediction unit 130 predicts that it will flow out from each inflow path at its own intersection. Calculated traffic flow (predicted outflow traffic flow). Specifically, the number of outflows in each traveling direction from each inflow path at its own intersection is calculated for each time t within the prediction target time range. That is, with reference to the signal control parameter data 341, the present at the prediction target time t is determined, and it is determined whether or not the right of passage is given to each inflow path according to the determined present.

  For the inflow path to which the right of passage is not given, as described with reference to FIG. 5, the number of staying in each lane of the inflow path is calculated. That is, each lane of the inflow path is determined according to the rate of progress in each traveling direction based on the number of vehicles that arrive at the inflow path at the predicted time t (arrival number) obtained by referring to the predicted arrival traffic flow data 332. The number of vehicles arriving at (the number of arrivals by lane) is calculated.

  The progress rate in each direction is stored in the progress rate table 322. FIG. 15 shows an example of the data configuration of the progress rate table 322. According to the figure, the progress rate table 322 stores a travel direction 322a and a progress rate 322b in association with each other. The traveling direction 322a has three directions of "straight forward", "left turn", and "right turn" because the self-intersection is a cross intersection. The progress rate 322b is determined so that the sum in all the traveling directions is “1.0”.

Next, the outflow traffic flow prediction unit 130 adds the calculated number of arrivals by lane to the number of stays at the time t ₋₁ immediately before the prediction time t for each lane of the inflow path, and stays in the lane at the prediction time t. Calculate the number.

  Here, the calculated staying number is stored in the staying number data 334 and the outflow number is stored in the outflow number data 335 for each inflow path.

  The staying number data 334 is data of the number of staying in each inflow path of the own intersection. FIG. 16 shows an example of the data configuration of the staying unit data 334. According to the figure, the staying number data 334 stores the staying number 334b of each lane in association with each time 334a within the prediction target time range.

  The number of outflows by inflow channel data 335 is data of the number of outflows from each inflow channel at the intersection in each traveling direction. FIG. 17 shows an example of the data configuration of the inflow channel outflow quantity data 335. According to the figure, the outflow number data 335 for each inflow path is generated for each inflow path at its own intersection, and the outflow number 335b from the inflow path in each traveling direction is calculated for each time 335a within the predicted time range. Stored in association. Although the data structure about the inflow path A is shown in the same figure, it is the same structure also about other inflow paths B-D.

On the other hand, as described with reference to FIG. 6, for the inflow path to which the right of passage is given, the number of staying in each lane of the inflow path and the number of outflows from the inflow path in each traveling direction are calculated. That is, in the same manner as the inflow path to which the right of passage is given, each number of the inflow path is determined according to the progress rate in each traveling direction based on the number of arrivals at the predicted time t obtained by referring to the predicted arrival traffic flow data 332. Calculate the number of arrivals to the lane (number of arrivals by lane). Next, for each lane in the inflow path, the number of arrivals by lane calculated at the time t- ₁ immediately before the prediction time t is added to calculate the number of temporarily staying lanes at the prediction time t. Subsequently, for each lane of the inflow path, it is determined whether or not the calculated temporary staying number of the lane exceeds the saturated traffic flow.

  The saturated traffic flow for each lane is stored in the saturated traffic flow table 323. FIG. 18 shows an example of the data configuration of the saturated traffic flow table 323. According to the figure, the saturated traffic flow table 323 stores a saturated traffic flow 323c in association with each lane 323b of each inflow channel 323a. The saturated traffic flow 323c is a value per lane.

  For a lane whose temporary staying number does not exceed the saturation traffic flow, the calculated temporary staying number is set as the number of outflows from the lane. Based on the number of outflows, the number of outflows from the lane in each direction of travel is calculated according to the rate of progress in each direction of travel, and the inflow at a time (t + Δt) after a predetermined intersection passage time Δt from the predicted time t. The number of spills from the road in the relevant direction of travel. Further, the staying number of the lane at the predicted time t is set to “0”.

On the other hand, for a lane in which the temporarily staying number exceeds the saturated traffic flow, this saturated traffic flow is set as the number of outflows from the lane. Then, based on the number of outflows, the number of outflows from the lane in each direction of travel is calculated according to the progress rate in each direction of travel, and the calculated number of outflows is calculated as a time ( The number of outflows from the inflow path to the corresponding traveling direction at t + Δt). Further, the number of outflows from the calculated lane is subtracted from the number of staying at the time t- ₁ immediately before the predicted time t of the lane, and the number of staying in the lane at the predicted time t is calculated.

  Furthermore, when calculating the number of outflows in each traveling direction from each inflow path, the calculated number of outflows is corrected according to the traffic volume of the traffic that intersects in the traveling direction (crossing traffic). That is, with reference to the crossing traffic table 324, the traffic crossing in the advancing direction of prediction object (crossing traffic) is determined.

  The crossing traffic table 324 is a data table that defines crossing traffic for each traveling direction from each inflow path. FIG. 19 shows an example of the data configuration of the crossing traffic table 324. According to the figure, the crossing traffic table 324 stores the crossing traffic 324c in association with each traveling direction 324b from each inflow path 324a.

  When the traveling direction to be predicted intersects with vehicle traffic, the number of outflows in the traveling direction is corrected according to the traffic volume of the vehicle traffic. That is, referring to the staying number data 334, the total number of staying numbers between the predicted time t and the past predetermined time (for example, 3 seconds) in the lane corresponding to the intersecting vehicle traffic is calculated. Then, if the calculated total number exceeds a predetermined threshold number (for example, five), the number of outflows in the traveling direction of the prediction target is changed to “0”.

  Moreover, when the advancing direction of a prediction object crosses with a pedestrian traffic, according to the traffic volume of the said pedestrian traffic, the outflow number to the said advancing direction is correct | amended. That is, with reference to the pedestrian crossing situation data 336, it is determined whether or not the intersecting pedestrian traffic is crossing at the predicted time t. If it is determined that the vehicle is crossing, the traffic volume in the traveling direction is changed to “0”.

  The pedestrian crossing situation data 336 is data of the pedestrian crossing situation at the pedestrian crossing of each inflow path of the own intersection. In FIG. 20, an example of a data structure of the pedestrian crossing situation data 336 is shown. According to the figure, the pedestrian crossing situation data 336 includes the number of pedestrians 336b, the crossing time 336c, the crossing start time 336d, and the crossing end time 336e for each of the inflow paths 336a provided with the crosswalk. Stored in association.

  If the predicted time t is a time between the crossing start time ts and the crossing end time te of the inflow path corresponding to the intersecting pedestrian traffic, it is determined that the pedestrian traffic is crossing, otherwise it is not crossing. Is determined.

  The pedestrian crossing situation data 336 is updated at the current switching timing predicted from the signal control parameter. That is, when the predicted time t is the current switching timing, for example, a predetermined coefficient is set for the number of stays for each inflow path to which the right of passage is given by the display after switching, in proportion to the number of stays in the inflow path. Crossing pedestrians are calculated by multiplication. Next, based on the calculated number of crossing pedestrians, a crossing time tu that is a time required for the crossing pedestrian to cross is calculated according to the crossing time setting data 325. The predicted time t, which is the current switching timing, is set as a crossing start time ts, and the time after the crossing time tu calculated from the crossing start time ts is set as a crossing end time te.

  The crossing time setting data 325 is data defining the relationship between the number of pedestrians and the crossing time tu, and is, for example, data of a functional expression of the graph shown in FIG.

Thus, for each inlet channel of the self intersection (from the current time t _0, until time n after a predetermined time) time range of the prediction target when calculating the travel direction by the outlet number at each time t of the generated flows The number of outflows from each inflow path is calculated based on the outflow number data 335 by road. That is, for each inflow path at the intersection, for each time t within the time range to be predicted, the total number of outflows from each other inflow path with the inflow path in the traveling direction is calculated and As the number of outflows, predicted outflow traffic flow data 331 for the own intersection is generated.

  The outflow traffic flow prediction unit 130 repeatedly calculates the predicted outflow traffic flow at predetermined time intervals. Then, the generated predicted outflow traffic flow data 331 is transmitted to each traffic signal control device 20 at each adjacent intersection.

  Returning to FIG. 10, the signal control unit 140 controls each traffic signal 30 at its own intersection according to the signal control parameter calculated by the control parameter calculation unit 120.

  The communication control unit 200 controls communication with other traffic signal control devices 20 and external devices such as the central management device 10. For example, the predicted outflow traffic flow data 331 transmitted from the traffic signal control device 20 at the adjacent intersection is received, or the predicted outflow traffic flow data 331 calculated by the outflow traffic flow prediction unit 130 is used as traffic signal control at the adjacent intersection. It transmits to each apparatus 20.

  The storage unit 300 stores a system program for the processing unit 100 to control the traffic signal control device 20 in an integrated manner, a program and data for realizing the traffic signal control of the present embodiment, and the processing unit 100. Used as a work area, and temporarily stores calculation results and the like executed by the processing unit 100 according to various programs. The storage unit 300 is realized by various IC memories, a hard disk, a ROM, a RAM, and the like, for example. In the present embodiment, the storage unit 300 includes a traffic signal control program 310, a self-intersection configuration table 321, a progress rate table 322, a saturated traffic flow table 323, a crossing traffic table 324, and crossing time setting data 325. , Predicted outflow traffic flow data 331, predicted arrival traffic flow data 332, temporarily staying number data 333, staying number data 334, outflow number outflow number data 335, pedestrian crossing situation data 336, signal control parameters Data 341 is stored.

[Process flow]
FIG. 21 is a flowchart for explaining the flow of the traffic signal control process. This processing is realized by the processing unit 100 executing the traffic signal control program 310 in the storage unit 300.

  According to the figure, first, the arrival traffic flow predicting unit 110 predicts the arrival traffic flow to the own intersection. That is, based on the predicted outflow traffic flow data 331 of the adjacent intersection received from the other traffic signal control device 20, the traffic flow (arrival traffic flow) arriving at each inflow path of the own intersection is predicted, and the predicted arrival traffic flow data. 332 is generated (step A1). A known method can be used for the prediction of the arrival traffic flow. Subsequently, the control parameter calculation unit 120 calculates the signal control parameter of the own intersection based on the generated predicted arrival traffic flow data 332 (step A3). This signal control parameter can also be calculated using a known method. Next, the outflow traffic flow prediction unit 130 executes the outflow traffic flow prediction process, predicts the traffic flow flowing out from each inflow path at the own intersection, and generates predicted outflow traffic flow data 331 (step A5). This outflow traffic flow prediction process is one of the characteristic processes of this embodiment.

FIG. 22 is a flowchart for explaining the flow of the outflow traffic flow prediction process. According to the figure, the outflow traffic flow prediction unit 130 first sets the current time t ₀ to the prediction time t (step B1). Then, a pedestrian crossing situation update process is performed (step B3).

  FIG. 23 is a flowchart for explaining the flow of the pedestrian crossing situation update process. According to the figure, the outflow traffic flow prediction unit 130 refers to the signal control parameter data 341 to determine whether or not the predicted time t is the current switching timing, and if it is the switching timing (step C1: YES), the process of loop D is performed for each inflow path at its own intersection.

  In loop D, it is determined whether or not a pedestrian crossing is provided in the target inflow path. If it is provided (step C3: YES), the right of passage to the pedestrian crossing the target inflow path according to the display after switching. Whether or not is given.

If the right of passage is given (step C5: YES), the pedestrian crossing the inflow path, for example, multiplying the number of stays by a predetermined coefficient based on the number of stays at the time t- ₁ immediately before the target inflow path. The number is calculated (step C7). Then, the crossing time tu is calculated based on the calculated number of crossing pedestrians (step C9), the predicted time t is set as the crossing start time ts, and the time obtained by adding the calculated crossing time tu to the crossing start time ts is calculated. The crossing end time te is set (step C11). Loop D is performed in this way. When the process of loop D for each inflow path at the own intersection is completed, the outflow traffic flow prediction unit 130 ends the pedestrian crossing situation update process.

  When the pedestrian crossing situation update process is completed, the outflow traffic flow prediction unit 130 performs the process of loop A for each inflow path of the own intersection. In loop A, based on the number of arrivals at the target inflow path at the predicted time t obtained by referring to the predicted arrival traffic flow data 332, the number of arrivals per lane of the target inflow path according to the progress rate in each traveling direction. Is calculated (step B5). Next, with reference to the signal control parameter data 341, the current indication at the predicted time t is determined, and it is determined whether or not the right of passage is given to the target inflow path (step B7).

  If the right of passage is given to the target inflow path (step B9: YES), the process of loop B is performed for each lane of the target inflow path.

In Loop B, the number of arrivals calculated is added to the number of stays at the time ₋₁ immediately before the target lane to calculate the temporarily staying number (step B11). Then, it is determined whether or not the calculated temporarily staying number exceeds the saturated traffic flow of the target lane. If not (step B13: NO), the temporarily staying number is set as the number of outflows from the target lane (step B15). . On the other hand, if the temporarily staying number does not exceed the saturated traffic flow (step B13: YES), this saturated traffic flow is set as the number of outflows from the target lane (step B17).

  Subsequently, loop C processing is performed for each of the traveling directions of the target lane. In the loop C, the number of outflows in the target traveling direction is calculated by multiplying the number of outflows from the target lane by the progress rate in the target traveling direction (step B19). Next, an outflow correction process is performed to correct the calculated number of outflows in the target traveling direction (step B21).

  FIG. 24 is a flowchart for explaining the flow of the outflow correction process. According to the figure, the outflow traffic flow prediction unit 130 determines whether or not there is traffic (crossing traffic) crossing in the target traveling direction. If there is crossing traffic (step D1: YES), the crossing traffic is determined. It is determined whether or not vehicle traffic is included.

  If vehicle traffic is included in the cross traffic (step D3: YES), the total number of temporarily staying vehicles in the past predetermined time from the predicted time t of the lane corresponding to the vehicle traffic that is the cross traffic is calculated (step D5). ). If the calculated total number exceeds the threshold number (step D7: YES), the number of outflows in the target traveling direction is changed to “0” (step D9).

  Moreover, if the pedestrian traffic is included in the intersection traffic in the target traveling direction (step D11: YES), the outflow traffic flow prediction unit 130 determines whether the pedestrian traffic that is the intersection traffic is crossing the pedestrian crossing. It is determined whether or not (step D13). If the vehicle is crossing (step D15: YES), the number of outflows in the target traveling direction is changed to “0” (step D17). When the above processing is performed, the outflow correction processing is terminated.

  When the outflow correction process ends, the outflow traffic flow prediction unit 130 determines the number of outflows in the target traveling direction from the target inflow path to the target traveling direction at a time (t + Δt) after a predetermined intersection passage time Δt from the predicted time t. (Step B23). Loop C is performed in this way.

  When the processing of loop C for each of the traveling directions of the target lane is completed, the outflow traffic flow prediction unit 130 subtracts the total number of outflows in each traveling direction from the temporarily staying number of the target lane, The number of staying at the predicted time t is calculated (step B25). Loop B is performed in this way.

On the other hand, if it is determined in step B9 that the right of passage is not given to the target inflow path (step B9: NO), the outflow traffic flow prediction unit 130 for each lane of the target inflow path immediately before the lane t. _The calculated number of arrivals at the lane is added to the number of staying at ₋₁ to calculate the number of staying at the predicted time t (step B27). In addition, the number of outflows from each lane in each traveling direction at time (t + Δt) after a predetermined intersection passage time Δt from the predicted time t is set to “0” (step B29). Loop A is performed in this way.

When the process of loop A for each inflow path at its own intersection is completed, the outflow traffic flow prediction unit 130 determines that the predicted time t is a time (t ₀ + T) after a predetermined elapsed time T from the current time t _0. If it has not been reached (step B31: NO), the predicted time is updated to a time obtained by adding a predetermined time (for example, 1 second) (step B33), and then the process returns to step B3. . On the other hand, if the predicted time t has reached the time (t ₀ + T) (step B31: YES), for each inflow path, add the number of outflows from other inflow paths with the inflow path as the traveling direction. The number of outflows from the inflow path is calculated, and predicted outflow traffic flow data 331 is generated (step B35). When the above processing is performed, the outflow traffic flow prediction processing is terminated.

  When the outflow traffic flow prediction process ends, the processing unit 100 transmits the generated predicted outflow traffic flow data 331 to each traffic signal control device 20 at each adjacent intersection (step A7). Then, it returns to step A1. The processing unit 100 repeatedly executes the series of processes of steps A1 to A7 at a predetermined time interval (for example, every 5 seconds).

[Action / Effect]
As described above, according to the present embodiment, when calculating the number of outflows in each traveling direction from each inflow path at the intersection, the number of outflows in each traveling direction depends on the traffic flow of the intersection traffic with respect to the traveling direction. It is corrected. Specifically, at the time of a right turn, the total number of oncoming straight vehicles from the predicted time t in the past predetermined time (that is, the total number of arrival and staying in the inflow path facing in the traveling direction) exceeds a predetermined threshold number. In this case, the number of outflows in the direction of the right turn is corrected to “0”, and the number of outflows in the direction of travel is set to “0” during the time corresponding to the pedestrian crossing the inflow path in the direction of travel when turning right or left. It is corrected. Thereby, the prediction of the outflow traffic flow with higher accuracy is realized.

[Modification]
The application of the present invention is not limited to the above-described embodiment, and it is needless to say that changes can be made as appropriate without departing from the spirit of the present invention.

(A) Correction of outflow number For example, in the above-described embodiment, the outflow number is set to “0” as the correction of outflow number according to the traffic volume of the crossing traffic, but may be reduced. . Specifically, when the correction is made in accordance with the oncoming straight vehicle, “0.0-1... Is proportional to the traffic volume of the oncoming straight vehicle that is the mixed traffic (that is, the number of temporarily staying on the opposite inflow path). A coefficient k of “0” is determined. Then, the number of outflows is corrected by multiplying the calculated number of outflows by this coefficient k. Similarly, when correcting according to the number of crossing pedestrians, a coefficient k is determined so as to be proportional to the number of crossing pedestrians that are cross traffic, and this coefficient is added to the number of outflows calculated while the crossing pedestrian is crossing. The number of outflows is corrected by multiplying by k.

(B) Shape of intersection and presenting method In the above-described embodiment, the intersection to be controlled is a cross intersection, but it can also be applied to intersections of other shapes such as a three-way intersection or a T-shaped intersection. Further, although the presenting method is the two presenting method, it is needless to say that the presenting method can be applied to a case where the presenting method is a multiple presenting method of three or more presenting methods.

Configuration of autonomous decentralized traffic signal control system. An example of an intersection. An example of presenting. Overview of signal control. The principle of outflow prediction for inflow channels without access rights. Principle of outflow prediction for inflow channels with right to pass. Explanation of correction of the number of outflows according to oncoming straight vehicles as cross traffic. Explanation of correction of the number of outflows according to crossing pedestrians as cross traffic. An example of the relationship between a crossing pedestrian and crossing time tu. The internal block diagram of a traffic signal control apparatus. The data structural example of a self-intersection structure table. Predicted outflow traffic flow data configuration example. Example of predicted arrival traffic flow data structure. The data structural example of signal control parameter data. The data structural example of a progress rate table. Data configuration example of staying unit data. Data configuration example of outflow data for each inflow channel. The data structural example of a saturated traffic flow table. The data structural example of a crossing traffic table. The data structural example of pedestrian crossing situation data. The flowchart of a traffic signal control process. The flowchart of the outflow traffic flow prediction process performed during a traffic signal control process. The flowchart of the pedestrian crossing situation update process performed during an outflow traffic flow prediction process. The flowchart of the outflow correction process performed during an outflow traffic flow prediction process.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Autonomous distributed traffic signal control system 10 Central management apparatus 20 Traffic signal control apparatus 100 Processing part 110 Arrival traffic flow prediction part, 120 Control parameter calculation part 130 Outflow traffic flow prediction part, 140 Signal control part 200 Communication control part 300 Storage part 310 Traffic Signal Control Program 321 Progress Rate Table, 322 Saturated Traffic Flow Table 323 Crossing Traffic Table, 324 Crossing Time Setting Data 331 Predicted Outflow Traffic Flow Data, 332 Predicted Arrival Traffic Flow Data 333 Temporarily Staying Number Data, 334 Staying Number Data 335 Inflow Number of outflow data by road, 336 Pedestrian crossing situation data 341 Signal control parameter 30 Traffic signal 40 Vehicle detector

Claims (7)

An autonomous decentralized traffic signal control device that transmits and receives information on the outflow traffic flow predicted with another traffic signal control device at another intersection,
Using the information of the outflow traffic flow from the other intersection received from the other traffic signal control device, the predicted value of the number of staying in each inflow path of the own intersection (hereinafter referred to as “predicted staying number”) and the own intersection A predicting means for predicting a predicted value of the number of outflows from each inflow path to the other path by outflow direction (hereinafter referred to as “predicted outflow number”) in time series at each prediction time ;
A means for correcting the predicted number of outflows from one inflow path to one outflow direction at one prediction time predicted by the prediction means, wherein traffic from the one inflow path to the one outflow direction is corrected. Predicted outflow number correcting means for predicting the traffic volume of the crossing traffic using the predicted staying number of the inflow path related to the crossing traffic, and performing the correction based on the predicted traffic volume,
Using the predicted outflow quantity corrected by changing the predicted outflow number by changing the one predicted time, the one inflow path, and the one outflow direction. A traffic signal control device that generates information and transmits the information to the other traffic signal control device. When the direction from the one inflow path to the one outflow direction is a right turn direction, the predicted outflow number correcting means sets the opposite straight vehicle and / or the opposite left turn car of the one inflow path as the crossing traffic. , Having a right turn outflow number correcting means for predicting the traffic volume of the crossing traffic using the predicted staying number of the opposite inflow path of the one inflow path, and performing the correction,
The traffic signal control device according to claim 1 . The right turn outflow number correction means calculates the traffic volume of the intersection traffic by using the predicted staying number of the past predetermined time from the one predicted time among the predicted staying number of the opposite inflow path of the one inflow path. Predict,
The traffic signal control device according to claim 2 . The predicted outflow quantity correcting means, when the direction of the from one inlet passage to the one outflow direction is a right turn direction, the crosswalk pedestrian outflow path of the outflow direction of the one and the crossing traffic, the Pedestrian reference right turn outflow number correction means for predicting the traffic volume of the crossing traffic using the predicted staying number of the opposite inflow path of one inflow path, and performing the correction,
The traffic signal control device according to any one of claims 1 to 3 . When the direction from the one inflow path to the one outflow direction is a left turn direction, the predicted outflow number correction means sets the crosswalk as a crosswalk pedestrian on the outflow path in the one outflow direction, Predicting the traffic volume of the crossing traffic using the predicted staying number of one inflow path, and having a pedestrian-based left turn outflow number correction means for performing the correction,
The traffic signal control apparatus as described in any one of Claims 1-4. The predicted outflow number correcting means reduces the predicted outflow number from the one inflow path at the one predicted time in the one outflow direction by a reduction rate corresponding to the traffic volume of the cross traffic. I do,
The traffic signal control device according to any one of claims 1 to 5. An outflow traffic flow prediction method in an autonomous decentralized traffic signal control apparatus that transmits and receives information on outflow traffic flow predicted with another traffic signal control apparatus at another intersection,
Using the information of the outflow traffic flow from the other intersection received from the other traffic signal control device, the estimated staying number of each inflow path of the own intersection , and the outflow direction from each inflow path of the own intersection to the other road A prediction step for predicting the number of predicted spills in time series for each prediction time ;
A step of correcting the predicted number of outflows from one inflow path to one outflow direction at one prediction time predicted in the prediction step, wherein traffic from the one inflow path to the one outflow direction is corrected. A predicted outflow quantity correction step of predicting the traffic volume of the crossing traffic using the predicted staying number of the inflow path related to the crossing traffic, and performing the correction based on the predicted traffic volume;
Using the predicted outflow number corrected by executing the predicted outflow number correcting step by changing the one predicted time, the one inflow path, and the one outflow direction, the flow of the outflow traffic from its own intersection An outflow traffic flow information generation step for generating information;
Outflow traffic flow prediction method including .

Images