JP2000242884A - Traffic flow simulation system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve prediction precision by bringing a traffic flow simulation close to the real flow. SOLUTION: When one node exists in an arrival mesh and the two nodes exist in a departure mesh, a departure node is selected in terms of probability by a formula Pj=(Rj-αJj)/ijΣn(Rj-αJi). Here, Rj is the weight coefficient of the node j(Nj), α is an adjustment parameter for optimizing the simulation system and n is 2. Besides, Jj is the congestion level of the node j, is 1 in the case of spill-back in a link connected to the node j and is 0 in the opposite case. When the departure node is selected by a probability Pj, the probability for selecting the node j as the departure node is changed by the congestion level Jj of the node j obtained by the simulation system so that congestion in the departure mesh is mitigated, and the simulation of the more realistic traffic flow is enabled.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system for simulating road traffic, and more particularly to a method for simulating road traffic by nodes corresponding to predetermined points on a road such as a signalized intersection and links corresponding to one side of the road between the nodes. The network is modeled, and the generated traffic volume, which is the traffic volume generated when the general vehicle running on the link starts to run, and the concentrated traffic volume, which is the traffic volume that disappears when the general vehicle arrives at the destination, are handled as nodes. The present invention relates to a traffic flow simulation system that simulates a traffic flow on a road network model by setting a traffic flow simulation model.

[0002]

2. Description of the Related Art Originating and terminating traffic (hereinafter referred to as “OD traffic data (Origin traffic)”
-Destination trip data) ”) can be estimated from socio-economic indicators such as census results. However, from the information of these survey results, although the OD traffic data between the respective sections in the predetermined area unit defined on the road network model can be estimated, it cannot be obtained as the OD traffic data between the respective nodes. Therefore, when simulating road traffic using the road network model as described above, it is necessary to convert the OD traffic data between the respective areas into OD traffic data between the nodes.

[0003] As a method of converting the OD traffic data between the predetermined areas into the OD traffic data between the nodes, for example, conventionally, an area having a relatively high population density according to the population distribution in each predetermined area is used. For example, a method of distributing a larger amount of generated / concentrated traffic is used for the node located at.

[0004]

However, such a conventional method of converting OD traffic data has the following drawbacks, so that a simulation close to an actual traffic flow (with high accuracy) is performed. It is difficult. (Disadvantage 1) The resident population, the working population, or the vehicle owner population in each region is not necessarily proportional to the vehicle driver population in each region directly related to the simulation. Further, the distribution of the driver population of the vehicle in each predetermined area may fluctuate depending on time. (Defect 2) Each vehicle does not always leave for the nearest node (intersection, etc.), and an advantageous link (road) under running conditions such as a large scale or a high speed limit tends to be more easily selected. is there. Accordingly, it is not always easy for the generated / concentrated traffic to easily concentrate on nodes located in an area having a relatively high population density. (Defect 3) Congested roads (nodes or links)
Since there is a tendency to avoid the departure point, the departure point (departure node) that is likely to be selected by each driver is sequentially different depending on road conditions.

The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to provide a traffic flow simulation system capable of accurately simulating an actual traffic flow.

[0006]

In order to solve the above-mentioned problems, the following means are effective. That is, the first means models a road network by nodes corresponding to a predetermined point on a road such as a signalized intersection, and links respectively corresponding to one side of the road between the nodes. The traffic flow on the road network model is set by setting the generated traffic volume, which is the traffic volume generated at the start of traveling, and the concentrated traffic volume, which is the traffic volume that disappears when a general vehicle arrives at the destination, in correspondence with the nodes. Traffic flow simulation system for simulating traffic flow, a generated / concentrated traffic volume prediction means for predicting a generated traffic volume and a concentrated traffic volume for each predetermined area defined on a road network model, Generating / centralized traffic distribution means for distributing the generated traffic volume for each predetermined area predicted by the means and the concentrated traffic volume for each node; By means, the nodes to which the generated traffic or the concentrated traffic should be distributed are determined by the weighting factor assigned to each node, the congestion degree of the node or the link connected to the node, or the direction or distance of the node to a predetermined area. , Stochastic,
Or, it is a deterministic choice.

[0007] The second means is the first means, wherein the weighting factor is determined by a road type of the link connected to the node or a predetermined minute time at an exit of the link connected to the node. It is determined by the saturated traffic flow rate, which is the maximum outflowable traffic volume. The road type of this link is highway, national road, main road, general prefectural road,
Although it can be classified into secondary arteries and other roads,
These may be further subdivided according to the number of lanes, road width, speed limit, and the like. With the above means, the above-mentioned problem can be solved.

[0008]

According to the above-mentioned means of the present invention,
The nodes to which the generated traffic or the concentrated traffic is to be distributed are stochastically determined by the weighting factor assigned to each node, the congestion degree of the node or the link connected to the node, or the direction or distance of the node to a predetermined area. Or deterministically. Therefore, since the departure / arrival node of the OD traffic data is determined in consideration of the weight of the nodes and the degree of congestion, the conversion of the OD traffic data depends only on the population density as in the prior art. Instead of conversion, a simulation based on actual traffic flow can be realized.
In particular, by the data conversion of the OD traffic data taking into account the current congestion obtained by simulation or the like, the previously obtained congestion is fed back to the OD traffic between the nodes. As a result, the OD between the nodes
Traffic data is more likely to be generated in a direction in which congestion at departure / arrival nodes is reduced. Therefore, the OD between each area
The conversion from the traffic data to the OD traffic data between the nodes is closer to the actual traffic flow. That is, according to the means of the present invention, these effects result in the above-mentioned (defect 1) to (defect 3) being eliminated, so that it is possible to more accurately simulate the actual traffic flow. Become.

[0009]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described based on specific embodiments. FIG. 1 is a hardware configuration diagram of a traffic flow simulation device 100 according to the present embodiment. The traffic flow simulation device 100 can be mounted on a car navigation system of a vehicle or the like. The CPU 101 uses the ROM 102 and the RAM 103 as storage devices, and executes various storages and calculations according to prepared programs. The CPU 101 inputs information necessary for specifying the current position and the traveling direction of the vehicle from the vehicle sensor 110 or the position information receiving device 109 via the input / output interface (IF) 108 as needed.

The display device 104 can display the travel time expected to be required to reach the destination. C
The PU 101 has a disk device (DK) 105, a ROM 1
02 and the information held in the RAM 103 constitute a database for storing link information of the road network model in link units. The CPU 101 controls the traffic information receiving means 10
6, traffic information is received in real time from a VICS or the like via an input / output interface (IF) 107 (traffic information input means).

FIGS. 2A and 2B are image diagrams of road network data in this embodiment. FIG. 2A is an image diagram when a road network model is displayed on the screen based on the road network data, and FIG. 2B is a configuration diagram of a table holding node data. , (C) is a configuration diagram of a table holding link data. In this way, the road network model can be logically constructed by holding the road network data on the database.

FIG. 3 is a schematic diagram of a road network model 300 according to an embodiment of the present invention. In this embodiment, as described above, nodes (N0, N1, N2,...) Corresponding to predetermined points on a road such as a signalized intersection, and links (L01, L12) respectively corresponding to one side direction of the road between these nodes. , L2
3)) to model the road network. So, for example,
For one-way roads, only one-way links are defined. Also, another link that shares the same node as a certain link and that can directly move (pass) traffic with the one link is called a connection link of the one link, and is called a connection link of the one link. That the upstream link,
The downstream connection link is called a downstream link.

FIG. 4 is an explanatory diagram showing the definition of the traffic volume of each link in the above road network model. For example, the link L12 of the road network model 300 (FIGS. 3 and 4)
Has three upstream links (L01, L41, L51) and three downstream links (L23, L26, L27). At this time, from all these upstream links, link L12
Is referred to as the inflow traffic volume (Q _in ) of the link L12. The accumulated value of the traffic flowing out from the link L12 to the downstream node N2 is referred to as the outflow traffic (Q _out ) of the link L12. Further, the traffic currently existing on the link L12 is referred to as the stopped traffic (Q _st ) of the link L12.

In the road network model of this embodiment, the generation of traffic volume due to the departure of a general vehicle (start of travel) and the disappearance of traffic volume due to the arrival of a general vehicle (end of travel) occur at each node. Shall be. For example, the cumulative value of the traffic volume generated at the node N1 and entering the link L12 is referred to as the generated traffic volume (Q _gen ) of the link L12. Also,
Node N2 of the outflow traffic volume (Q _out ) of link L12
Centralized traffic volume of the link L12 disappearance to traffic in (Q _a)
Say The above four traffic volumes Q _in , Q _out , Q _st , Q
_Gen is generally referred to as link traffic. This link traffic volume is information included in the link information.

FIG. 5 is a functional block diagram of the traffic flow simulation function in the present embodiment. The traffic flow simulation function 50 includes a vehicle departure destination point determination function 51.
And a vehicle moving function 52 and the like. The vehicle departure destination point determination function 51 is provided for each mesh between predetermined areas.
The D traffic data 55 is converted into OD traffic data 57 of a general vehicle between nodes. In performing this data conversion, the road network data 56 and the congestion state 58 of each road, which is the simulation result, are used as selection criteria for departure / arrival nodes.

The mesh-specific OD traffic data 55 includes a departure time (traffic generation time) of each vehicle, a departure area and a destination area assigned to a mesh (predetermined area). The road network data 56 is the data shown in FIG. 2, and the node data includes a weight coefficient of the node. OD traffic data 57 of general vehicles between nodes
Is a data conversion result of the vehicle departure destination point determination function 51, and has departure point (departure node) information and destination point (arrival node) information of each vehicle. Congestion situation 58 of each road
Is a simulation result of the vehicle movement function 52.

The vehicle movement function 52 executes a vehicle movement simulation based on the OD traffic data 57 of the ordinary vehicle between the nodes and the road network data 56, and congests each road or a designated departure point. -Predict travel time between destinations (prediction result 53 in FIG. 5).

FIG. 6 is a diagram showing the configuration of a table for storing (a) mesh-specific OD traffic data 55 in the present embodiment;
(B) is a configuration diagram of a table that holds vehicle occurrence time data 60. The vehicle occurrence time data 60 is data generated by the vehicle departure destination point determination function 51 as intermediate processing data, and based on the mesh-specific OD traffic data 55, uses a random number to determine when each general vehicle departs. It is determined stochastically.

FIG. 7 is a flowchart showing a procedure for simulating a road network model in the present embodiment. This flowchart is executed by the CPU 101 (FIG. 1). In this flowchart, first, at step 700, the vehicle occurrence time data 60 is generated. Next, at step 710, the stopped traffic volume (Q _st ) is set for all the links. This setting is performed based on reference values for each day of the week and for each time zone stored in a link traffic volume table (not shown).

In step 720, a processing completion target time T (= t + ΔT) is set from the current time (timer time) t, the processing step t _{s of} this simulation, and the processing target value ΔT. In step 730, in the vehicle occurrence time data 60 generated in step 700, unprocessed data whose occurrence time has reached the current time t (vehicle generation processing 76
0 is not implemented)
If there is unprocessed data, the process proceeds to step 740; otherwise, the process proceeds to step 770.

In step 740, a departure / arrival node is selected by a method described later. The most important feature of the present invention lies in the method of selecting the departure / arrival node.
In step 750, a traveling route between the departure and arrival nodes of the corresponding general vehicle is determined by the Dijkstra method or the like. In step 760, a vehicle generation process is performed. By this processing, the generated traffic volume (Q _gen ) of the corresponding link is updated, and the vehicle movement processing in step 770 becomes possible.

In step 770, the moving process of each link of the general vehicle is performed by the following model, the block density method, and the like. This movement processing, the traffic volume Q _in of each link on the road network model, Q _out, such as the Q _st is updated. Here, the passing time of each link can be obtained from the time when the vehicle enters and exits the link. In step 780, it is determined whether or not the current time t has reached the processing completion target time T.
If not, go to step 730.

By the action of this step, step 73
Since a series of processes from 0 to step 780 is repeatedly executed, data conversion of OD traffic data taking into account the current congestion degree on each link obtained by simulation becomes possible. That is, the congestion degree obtained earlier is
In step 740, the OD traffic data between the nodes newly obtained is fed back. Thereby, the OD traffic data between the nodes is easily generated in a direction in which the congestion at the departure / arrival nodes is reduced. Therefore, by this operation, the conversion from the OD traffic data between the areas to the OD traffic data between the nodes becomes closer to the actual traffic flow.

In step 790, the link transit time of each link or the traffic congestion state of each link is obtained from the link information including the link traffic volume calculated and updated by the above processing. Through the above processing, the link transit time of each link on one route from the departure point to the destination point or the traffic jam occurrence state at each link can be accurately obtained.

FIG. 8 shows vehicle occurrence time data 60 (FIG. 6).
(B) and FIG. 8 (a)), the OD of the general vehicle between the nodes
FIG. 9 is an explanatory diagram showing a data conversion method for generating traffic volume data 57 (FIGS. 5 and 8B). This data conversion processing is performed in step 740 described above.

FIG. 8 (a) shows the above-mentioned step 70 based on the mesh-specific OD traffic data 55 between the predetermined areas.
This is a schematic representation of an image of the vehicle occurrence time data 60 generated by using 0. That is, in FIG. 8A, the departure mesh (departure area) and the arrival mesh (arrival area) of a certain general vehicle are determined, and the current time t is the departure time of this vehicle (traffic generation time). ) Is reached.

At this time, in the simulation system, a departure node and an arrival node on the road network model of the general vehicle are specifically determined. For example, FIG.
As shown in (1), when there is only one node (node 3) in the arrival mesh and two nodes (node 1 and node 2) in the departure mesh, the departure node is represented by the following equation ( It is selected stochastically by the probability P _j of 2).

[Number 1] _{P j = (R j -αJ j} ) / i = 1 Σ n (R i -αJ i) ... (2)

Here, R _j is a weight coefficient of the node j (Nj) shown in FIG. 2B, α is an adjustment parameter (positive real number) for optimizing the simulation system, and n = 2. J _j is the congestion level of node j, and is 1 when spillback occurs on the link connected to node j, and 0 when it is not.

When the starting node is selected based on the probability P _j , the probability that the node j is selected as the starting node changes depending on the congestion level J _j of the node j obtained by the simulation system. The congestion in the departure mesh is reduced, and a more realistic simulation of traffic flow can be performed. In addition, such a node selection method can be used in the same manner as described above when determining an arrival node even when a plurality of arrival nodes exist in the arrival mesh.

The selection of a node may be deterministic. For example, the node existing in the arrival mesh is 1
If there is no node in the starting mesh and no node exists in the starting mesh, in this embodiment, the node having the largest value of A _j defined by the following equation (3) is started from all the nodes. Select as a node.

A _j = R _j −αJ _j + β cos θ _j + ν / r _j ⁿ (α> 0, β> 0, ν> 0, n> 1) (3) where β and ν are Tuning parameters (positive real numbers) for optimizing the simulation system, and (r _j , θ _j ) is the starting node j (N
j) are music coordinates.

FIG. 9 is an explanatory diagram of the music coordinates used when performing data conversion from vehicle occurrence time data to OD traffic data between nodes in the present embodiment. In the figure, the point O is the center point of the starting mesh and the origin of the curved coordinates (r _j , θ _j ). Point C is the center point of the arrival mesh.

If a node having the largest A _j shown in (3) is selected using such a piece of music coordinates, a node located in a direction closer to the direction of the destination or a center point closer to the departure mesh is selected. node at the same time is easily more selective, large weighting coefficient R _j nodes, it becomes easy to be selected as the starting node as fewer nodes of congestion, congestion is likely to be relaxed, and can simulate more realistic traffic flow Become. Also, such a node selection method can be used in the same manner as described above when determining an arrival node.

According to the present invention, since the departure / arrival node can be determined even when no node exists in the mesh, the size of the mesh can be easily changed. There is also an effect. Alternatively, even if the number of links and the number of nodes of the road network model are changed, the inter-node OD traffic data 57 is obtained by the operations of steps 700 and 740 based on the mesh-specific OD traffic data 55 between the predetermined areas. Since it is generated automatically and dynamically (in real time), there is an effect that a great degree of freedom is generated in the configuration of the road network model. That is, according to the present invention, since the OD traffic data is allocated according to the road network model (road network data), the OD traffic data can be used regardless of the change of the network.

[0034] In the above embodiment, although the congestion level J _j weighting factors R _j and node j of the node is using discrete numerical values, the weighting factors of nodes R _j and node j congestion level J _j may be a continuous real number. That is,
In addition to the road type of the link connected to the node, the weighting coefficient may be, for example, a saturated traffic flow rate (continuous real number), which is the maximum possible outflow traffic per minute minute at the exit of the link connected to the node. ) Can be determined.

The road type of this link can be classified into expressways, national roads, main roads, general prefectural roads, secondary trunk roads, other roads, and the like. These are further divided into the number of lanes, road width, and speed limit. It may be subdivided by such as.

In the above embodiment, the arrival node and the traveling route of the general vehicle are determined by the vehicle movement processing (step 7).
Although it has been determined before the execution of step 70), the arrival node and the traveling route of the general vehicle are determined by the vehicle movement processing (step 77
0) may be determined at the same time as the execution, and for example, a method of stochastically determining with time may be used. Even in such a case, the present invention exerts its effect sufficiently at the stage of determining a certain starting node from a certain starting mesh.

Further, in this embodiment, a traffic flow simulation device which can be mounted on a car navigation system of a vehicle has been specifically described. However, this device can be applied to a traffic control center, a taxi company, a transportation company, and other command centers. It can also be used as a traffic flow simulation device installed in

[Brief description of the drawings]

FIG. 1 is a hardware configuration diagram of a traffic flow simulation device according to an embodiment of the present invention.

FIG. 2 is an image diagram of road network data in the embodiment of the present invention.

FIG. 3 is a schematic diagram of a road network model according to the embodiment of the present invention.

FIG. 4 is an explanatory diagram showing a definition of a link traffic volume in the embodiment of the present invention.

FIG. 5 is a functional block diagram of a traffic flow simulation function according to the embodiment of the present invention.

FIG. 6 shows (a) OD for each mesh in the embodiment of the present invention.
Configuration diagram of a table that holds traffic volume data, and
(B) Configuration diagram of a table that stores vehicle occurrence time data.

FIG. 7 is a flowchart showing a procedure of simulating a road network model in the embodiment of the present invention.

FIG. 8 is an explanatory diagram showing a data conversion method from vehicle occurrence time data to OD traffic data between nodes in the embodiment of the present invention.

FIG. 9 is an explanatory diagram of music coordinates used when performing data conversion from vehicle occurrence time data to OD traffic data between nodes in the embodiment of the present invention.

[Explanation of symbols]

IF ... output interface DK ... disk apparatus Nm ... node number (m is an integer) Ll ... link number (l is an integer) tau ... link passing time Q _in ... inflow traffic volume (cumulative value) Q _out ... outflow traffic (cumulative Value) Q _gen … Generated traffic volume (cumulative value) Q _a … Centralized traffic volume (cumulative value) Q _st … Stopped traffic volume (current value) Q _lim … Saturated traffic flow rate (maximum possible outflow traffic per unit time) R _j … node weighting factor J _j … congestion level of node j (Nj) α… optimization adjustment parameter (positive real number) β… optimization adjustment parameter (positive real number) r _j … node j ( Nj) tuned coordinates (distance from origin O) θ _j ... tuned coordinates of node j (Nj) (angle from positive x-axis direction)

Claims (2)

[Claims] 1. A road network is modeled by nodes corresponding to predetermined points on a road such as a signalized intersection, and links respectively corresponding to one side of the road between the nodes, and traveling of general vehicles traveling on the links. The traffic volume on the road network model is set by setting the generated traffic volume, which is the traffic volume generated by the start of traffic, and the concentrated traffic volume, which is the traffic volume that disappears when the general vehicle arrives at the destination, in correspondence with the node. In a traffic flow simulation system for simulating a flow, the generated traffic volume and the concentrated traffic volume prediction means for predicting the generated traffic volume and the concentrated traffic volume for each predetermined area defined on the road network model; The generated traffic volume of the predetermined area unit predicted by the concentrated traffic volume prediction means, and the generated / concentrated traffic volume for distributing the concentrated traffic volume corresponding to the node. Means, wherein the generated / concentrated traffic distribution means is connected to the node or the node, wherein the node to which the generated traffic or the concentrated traffic is to be distributed is assigned a weighting factor assigned to the node. A traffic flow simulation system, which selects stochastically or deterministically according to the degree of congestion of the link or the direction or distance of the node to the predetermined area. 2. The saturation factor, wherein the weighting factor is a road type of the link connected to the node, or a maximum possible outflow traffic amount per predetermined minute time at an exit of the link connected to the node. The traffic flow simulation system according to claim 1, wherein the traffic flow simulation system is determined by a traffic flow rate.