JP5491104B2 - Traffic condition analysis device, traffic condition analysis program, and traffic condition analysis method - Google Patents

Description

  The present invention relates to a traffic situation analysis device, a traffic situation analysis program, and a traffic situation analysis method for analyzing a traffic situation of a road from information relating to a traffic situation of a vehicle.

When providing traffic information, the traffic condition analysis device collects road information data related to the traffic condition of the vehicle, determines the traffic condition based on the road information data, and provides the determined traffic information Used when.
The road information data of each place is regularly collected by the traffic information system, and the collected information becomes basic data.
The traffic information provided so far includes a VICS (Vehicle Information and Communication System) that uses an in-vehicle car navigation device as a receiver, and a broadcast media (for example, Non-Patent Document 1). reference). Information provided by VICS is also provided by a distribution server connected to the Internet, and can be acquired in response to a request from a terminal device or a mobile phone terminal device connected to the Internet.

On the other hand, the traffic information through the conventional broadcast media is provided under the following restrictions. For example, the target routes on which information is provided are limited to highways and main highways. In other words, traffic information on a limited number of target routes is provided, but information on the status of roads other than the main trunk road is not provided (for example, see Non-Patent Document 2).
In addition, the display of the provided information is limited in expressive power limited by the resolution that can be displayed on the television monitor. Therefore, it is provided by a method in which the map is schematically shown to show the target route, and the display color of the target route is changed to display the congestion state.

Road Traffic Information and Communication System Center, http://www.vics.or.jp/area/tokyo.html (searched July 23, 2009) Japan Road Traffic Information Center, http://www.jartic.or.jp/traffic/ippan/tokyoseibu.html (searched July 23, 2009)

  By the way, the traffic information data so far includes data detected by traffic counters, etc. that are fixedly arranged along the main trunk roads, etc., and the traffic volume and speed of the vehicle at the location where the traffic counter is located Based on. 2. Description of the Related Art In recent years, it has been studied to detect a traveling state of a vehicle that is actually traveling and use it as provided information, considering the traveling vehicle as a moving sensor. Such a system is called "probe traffic information" and is expected to expand the scope of provision compared to traffic information data based on information detected by a fixed traffic counter. Yes.

However, as shown in Non-Patent Documents 1 and 2, information provision in broadcasting is intended to provide information for an unspecified number of people, and therefore, from information providers to information users Since it is based on a one-way information transmission form, there is a restriction on the information transmission direction.
In other words, in broadcasting information provision, it is not possible to display the direction weighted according to the direction of movement as in the car navigation system on the vehicle, and it is also necessary to operate like a mobile phone terminal. Necessary information cannot be requested, and it is required to refer to the broadcast information display. That is, in providing information via broadcasting media, it is necessary to provide a wide range because the target area of the information to be displayed cannot be specified. The conventional providing methods have a problem that it is impossible to provide information other than the main trunk road while providing a wide range.

  The present invention has been made to solve the above problems, and its purpose is not limited to information on main arterial roads and the like, but is provided with a traffic situation analysis device, a traffic situation analysis program, and a traffic situation analysis method. Is to provide.

In order to solve the above problem, the present invention provides a relationship between the number of vehicles existing in the area and the area traffic state information indicating the number of traveling vehicles passing through the area for each predetermined area unit. An approximate curve calculation unit that calculates an approximate curve indicated by a function that is convex upward, a congestion index calculation unit that calculates a congestion index indicating the degree of difficulty in passing the area based on the positional relationship with the approximate curve, A traffic condition analysis apparatus comprising: a change index calculation unit that calculates a change index from a difference between an approximate curve represented by an upward convex function as a reference and the area traffic state information .

Further, in the present invention according to the above invention, the approximate curve calculation unit calculates a parameter of a two-dimensional normal distribution by performing projective transformation along the approximate curve indicated by the upward convex function. Features.

  Further, in the present invention according to the above invention, the approximate curve calculation unit is a total indicated by the total travel time of a vehicle traveling per unit time on the road selected as the target road in the area for the area traffic state information. It is characterized by traffic density.

  Also, in the present invention according to the above-described invention, the approximate curve calculation unit is a total indicated by the total travel distance of a vehicle traveling per unit time on the road selected as the target road in the area. It is characterized by traffic volume.

  Further, the present invention is characterized in that, in the above invention, a congestion index output unit that outputs the change of the congestion index or the change index for each area in a stepwise index.

  Further, the present invention is the above invention, wherein the congestion index output unit maps the area on a map, and is based on a congestion index derived for each area or a stepwise index derived from the change index. The display is performed according to the difference in color or density of the area.

Further, in the present invention, for each predetermined area unit, the relationship between the number of vehicles existing in the area and the area traffic state information indicating the number of vehicles traveling through the area is convex upward. A procedure for approximating with an approximate curve indicated by a function, a procedure for calculating a congestion index indicating the difficulty of passing in the area based on the positional relationship with the approximate curve, and a function that is convex upward as a reference A traffic condition analysis program that causes a computer to execute a procedure for calculating a change index from the difference between the approximate curve and the area traffic condition information .

Further, according to the present invention, the approximate curve calculation unit has a relationship between the number of vehicles existing in the area and the area traffic state information indicating the number of traveling vehicles passing through the area for each predetermined area unit. a step of calculating a step of approximating an approximate curve shown in function which is convex upward, congestion index calculation unit, a congestion index that indicates the passage difficulty of the area based on the positional relationship between the approximate curve, A change index calculating unit, comprising: calculating a change index from a difference between the approximate curve indicated by the upward convex function as a reference and the area traffic state information; It is.

According to the present invention, in the traffic condition analysis device, the approximate curve calculation unit indicates the number of vehicles existing in the area and the number of traveling vehicles passing through the area for each predetermined area unit. An approximate curve indicating the relationship of the area traffic state information as a convex function is calculated. The congestion index calculation unit calculates a congestion index indicating the degree of difficulty in passing the area based on the positional relationship with the approximate curve.
Thereby, the traffic condition analysis apparatus can provide the traffic condition without being limited to information on the main trunk roads and the like.

It is a schematic block diagram of the traffic condition analysis apparatus by this embodiment by one Embodiment of this invention. It is a figure shown about the area traffic flow state in one road by this embodiment. It is a figure shown about the area traffic flow state in the road network by this embodiment. It is a figure which shows the unusualness of the traffic state by this embodiment. It is a figure which maps and shows distribution of the congestion degree by this embodiment on a map. It is a figure which maps and distributes the distribution of the traffic state by this embodiment on a map. It is a figure which shows the relationship between the area and road by this embodiment. It is a figure which shows total QK for every area by this embodiment. It is a figure which shows the total QK approximation curve for every area from the scatter diagram shown in FIG. 8 by this embodiment. It is a figure which shows the conversion from the quadratic function by this embodiment to parametric space. The distribution of the congestion index on weekdays and holidays according to this embodiment is shown respectively. It is a flowchart which shows the procedure of the initialization process in the procedure of the traffic condition analysis process by this embodiment. It is a flowchart which shows the procedure of the analysis process in the procedure of the traffic condition analysis process by this embodiment. It is a flowchart which shows the procedure of the approximate curve calculation process in the procedure of the traffic condition analysis process by this embodiment.

An embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a schematic configuration diagram of a traffic condition analysis apparatus according to the present embodiment.
The traffic condition analysis apparatus 100 shown in this figure periodically acquires the traffic condition data provided from the traffic information system 200 and analyzes the traffic condition at that time.
The traffic information system 200 periodically collects traffic condition data to be provided at a predetermined cycle. The traffic situation data provided by the traffic information system 200 is information indicating the traffic situation of the road included in the analysis target range, and detects, for example, information collected by a traffic counter arranged on the roadside or a running situation. This is information including information notified based on an actual traveling state detected by a vehicle having a function (probe car).

  The traffic situation analysis device 100 includes an aggregation target area information acquisition unit 111, an area information storage unit 112, an information acquisition unit 121, an area aggregation QK state value calculation unit 122, a deviation amount calculation unit 123, a change index calculation unit 124, and a congestion index output. Unit 125, area total QK state value storage unit 131, total QK approximate curve unit 132, and two-dimensional normal distribution parameter conversion unit 133.

The aggregation target area information acquisition unit 111 acquires information that defines a target area (range) for analysis processing performed by the traffic situation analysis device 100 and records the information in the area information storage unit 112. The information acquired by the tabulation target area information acquisition unit 111 is map information of the target range, route information, area information that defines a target area indicating a unit for performing traffic condition analysis, and information indicating a relationship between the target area and the route. is there. For example, the target area is defined by classification based on a predetermined distance or a range divided by latitude and longitude.
Moreover, the aggregation target area information acquisition unit 111 associates each target area with a route related to the area. Details of the association method will be described later.

  The area information storage unit 112 includes map information of the target range recorded by the total target area information acquisition unit 111, route information, area information for defining a target area indicating a unit for performing traffic condition analysis, and information on the target area and the route. Information indicating the association is recorded.

The information acquisition unit 121 periodically acquires traffic situation data provided from the traffic information system 200 at a predetermined cycle. For example, the information acquisition unit 121 acquires traffic situation data at a period set from several minutes to about 20 minutes. If this acquisition cycle is shortened, each computation processing amount to be described later increases, and if it is lengthened, detection of an event change is delayed, and notification of an analysis result that is computed based on the result is also delayed. Therefore, this acquisition cycle is determined in advance according to the time required for the applied system.
In addition, the information acquisition unit 121 includes map information of the target range stored in the area information storage unit 112, route information, information that defines a target area indicating a unit for performing traffic condition analysis, and correspondence between the target area and the route. Information relating to the range to be analyzed is acquired.
And the information acquisition part 121 acquires the number of vehicles which pass a target area, and the required time information concerning passage based on said information.

Based on the information acquired by the information acquisition unit 121, the area total QK state value calculation unit 122 calculates an area total QK state value indicating the traffic situation for each target area. Details of the method of calculating the area total QK state value will be described later.
The area total QK state value calculation unit 122 records the area total QK state value calculated for each target area in the area total QK state value storage unit 131. The recording in the area total QK state value storage unit 131 is performed in accordance with the calculation of the area total QK state value, and is performed in time series with time information corresponding to the information acquired by the information acquisition unit 121 being added. .

The divergence amount calculation unit 123 performs projective transformation on the area total QK state value calculated by the area total QK state value calculation unit 122 with respect to the total QK approximate curve derived based on the area total QK state value. Also, the divergence amount calculation unit 123 calculates the divergence amount from the area congestion index and the approximate curve from the result of the projective transformation.
The change index calculation unit 124 determines the degree of change from the two-dimensional normal distribution derived by the two-dimensional normal distribution parameter conversion unit 133, the area congestion index calculated by the deviation amount calculation unit 123, and the deviation amount from the approximate curve. A change index indicating the amount of entropy information indicating is calculated.
The congestion index output unit 125 outputs the calculated congestion index and change index for each area in accordance with the output form set in stages.

The area total QK state value storage unit 131 stores the area total QK state value calculated for each area by the area total QK state value calculation unit 122 according to time series.
The total QK approximate curve unit 132 refers to the area total QK state value for each area stored by the area total QK state value storage unit 131 and calculates a total QK approximate curve for each area. Details of the calculation method of the aggregate QK approximate curve will be described later.
The two-dimensional normal distribution parameter conversion unit 133 performs projective transformation of the area total QK state value stored in the area total QK state value storage unit 131 along the total QK approximate curve for each time zone, and generates a two-dimensional normal distribution. Calculate the parameters.

Next, the area traffic flow state will be described.
FIG. 2 is a diagram showing an area traffic flow state on one road.
FIG. 2A is a plan view showing a model of a state where a vehicle is passing on a two-lane road. A vehicle entering from the left side of the road is kept in the inter-vehicle distance and is in an easy-to-run state. In the state shown on the right half of the road, there is a state in which a traffic jam occurs, the vehicle interval is narrowed, and the vehicle speed is reduced. Further, when the vehicle interval becomes narrower, the vehicle shifts to a stopped state and the traffic volume is extremely reduced. That is, when the vehicle interval is sufficiently maintained, the traffic volume increases according to the number of vehicles. On the other hand, when shifting to a state where the vehicle interval is not sufficiently maintained, the traffic volume decreases according to the number of vehicles.

FIG. 2B is a graph showing the relationship between the density of vehicles on the road and the traffic volume. In this graph, the vertical axis indicates the traffic volume, that is, the number of passing vehicles per unit time, and the horizontal axis indicates the vehicle density, that is, the number of traveling vehicles per unit distance.
Until the maximum performance of this road, that is, the capacity that shows the maximum traffic volume, the higher the density of vehicles, the higher the traffic volume. In this state, there will be no traffic jams and the vehicle will be able to run freely. Indicates the state. On the other hand, when a certain density is exceeded, traffic congestion occurs, the number of vehicles per unit distance increases, and the speed decreases, thereby reducing the traffic volume.
The slope of the line segment connecting the origin and any point on the graph (for example, point A) shown in this graph indicates the average speed in a state where traffic congestion has occurred. When the number of vehicles further increases in a traffic jam, the speed decreases and the traffic volume decreases as the density increases. For this reason, the inclination indicated by the inclination of the line segment becomes small, and the average speed decreases.

  The states shown in FIGS. 2A and 2B show an ideal state. On actual roads, there are parked vehicles, which occupy the road space without running, causing a decrease in traffic. In addition, there is a signal and a time during which the movement of the vehicle is restricted is a factor that reduces the traffic volume. Furthermore, there is a restriction with pedestrians and bicycles, and a state where the vehicle cannot run freely occurs and the average speed is lowered, which causes a decrease in the traffic volume. In addition to these factors, since an event that causes a difference from the ideal state occurs on an actual road, it appears as a limit value with a traffic volume smaller than the traffic volume shown in FIG.

FIG. 3 is a diagram showing an area traffic flow state in the road network.
FIG. 3A is a plan view showing a road network arranged in a lattice pattern. Each road shown in this figure is a one-way lane facing road.
In the road network shown in this figure, when congestion occurs in some roads, the fluidity of the roads decreases, and traffic that tries to pass through other roads is obstructed. For this reason, the vehicle cannot flow into the congested road, and the vehicle that is about to flow in waits for the inflow before the intersection, and a road with few vehicles and no traffic jam occurs in front of the intersection.
Since there is a possibility that the liquidity of the entire area is reduced due to traffic congestion on some roads without using the performance of such roads to the limit, the liquidity of the entire area is included in that area It becomes smaller than the case where all roads are used up to the limit performance.

FIG. 3B is a graph showing the relationship between the number of traveling vehicles and the aggregate traffic volume of the traveling vehicles by extracting the traveling vehicles from the vehicles in the area.
In this graph, the vertical axis represents the total traffic volume, and the horizontal axis represents the number of all traveling vehicles in the area. The total traffic volume shown on the vertical axis indicates the total traffic volume of all roads in the area.
A graph G20 shows the limit performance of the road when the road is used in an ideal state, and a graph 21 shows the limit performance based on the actual use condition of the road.

The graph G20 in FIG. 3B also shows a tendency similar to the traffic volume shown by the graph 10 shown in FIG. 2B, and shows the maximum performance of these roads, that is, the maximum traffic volume. Until the capacity is reached, the total traffic volume increases as the number of traveling vehicles increases. In this state, traffic congestion does not occur, and a state in which the vehicle can freely travel is maintained. On the other hand, if the number of vehicles exceeds a certain number, traffic congestion occurs, the number of vehicles per unit distance increases, and the speed decreases, so the total traffic volume decreases.
Further, since the graph G21 sums up values based on actual road usage conditions, the graph G21 shows a lower value than the ideal graph G20 in any number of traveling vehicles. Further, the number of traveling vehicles showing the peak of the graph G21 is smaller than the number of traveling vehicles showing the peak of the graph G20.

Next, traffic state variability will be described.
FIG. 4 is a diagram illustrating the unusualness of the traffic state.
FIG. 4 shows the change in traffic observed on a three-lane highway.
In this graph, the vertical axis indicates the traffic volume (pcu / hour), that is, the number of passing vehicles per unit time, and the horizontal axis indicates the density (pcu / km), that is, the number of traveling vehicles per unit distance.
The graph of FIG. 4 shown in the scatter diagram corresponds to the graph of FIG.
The traffic tends to increase monotonically until the density reaches 75 (pcu / km). In addition, the traffic volume tends to decrease when the density exceeds 75 (pcu / km), and the variation becomes large. And in the density range from 75 (pcu / km) to 150 (pcu / km), the distribution of plots is dense, indicating that congestion in this density range occurs on a daily basis. .
Furthermore, when the density increases and exceeds 150 (pcu / km), the plot distribution becomes coarse, and the frequency of occurrence is reduced, but there are occasional traffic jams, so extraordinary traffic jams have occurred. It is shown.

Next, the geographical distribution of the congestion level will be described.
FIG. 5 is a diagram showing the congestion degree distribution mapped to a map.
The traffic situation of each area is judged at each stage of “no congestion”, “slightly crowded”, “crowded” and “traffic jam” for each unit indicated by the grid, and is displayed at a concentration corresponding to each stage. . This figure shows that the higher the concentration of each area, the lower the fluidity and the traffic congestion.
As shown in this figure, if the adjacent areas are at the congestion level indicated by the same level of density, it is indicated that the related traffic congestion has occurred across the areas. By dividing this congestion level into areas, it is possible to confirm the spread of traffic congestion.

The geographical distribution of traffic conditions will be described.
FIG. 6 is a diagram showing the traffic state distribution mapped to a map.
The traffic situation of each area is judged at each stage of “normal state”, “slightly more crowded than normal”, and “crowded more than normal” for each unit indicated by the grid, and displayed at a density corresponding to each stage. . In this figure, it is shown that as the concentration in each area is higher, the fluidity is lower than usual and traffic congestion occurs.
As shown in this figure, if the traffic situation is shown at the same level of concentration as the adjacent area, the related traffic congestion occurs across the area, and the traffic congestion is lower than usual. Is shown. In addition, a traffic jam related to the traffic jam occurs around the area where the traffic jam having a higher congestion level occurs.
By referring to the traffic state display shown in FIG. 6 in combination with the above-described congestion level display (FIG. 5), road users who are aware of the normal traffic condition can compare with the normal traffic condition. Can be easily performed.

Next, the relationship between the area into which the target range is divided and the road will be described.
FIG. 7 is a diagram illustrating the relationship between areas and roads. In this figure, the area indicated by a square indicates, for example, a range obtained by dividing the vertical and horizontal sides into (10 × 10) for a secondary mesh in a digital map, and is a regional mesh (tertiary mesh) divided by 1 km (km) square. Is the aggregation unit.
A node N1 located outside the area (regional mesh) and nodes N2 to N5 located inside the area are shown in FIG. Each node indicates a dividing point when defining a road connecting the nodes.

Nodes N2, N3, N4 and N5 in the area are connected by links L23, L34 and L45 indicating roads. Further, the node N1 and the nodes N2 and N3 in the area are connected by links L12 and L13, respectively.
Each link is defined with a direction. When nodes at both ends are included in this area, each link connecting the nodes is defined as belonging to this area in both directions, and when one node is included in this area, it is downstream It is defined that the link in the direction including the node belongs to this area. That is, a vehicle that moves in the area and a vehicle that flows into the area are defined as target vehicles in this area.

In some cases, expressways, motorways, etc. (hereinafter collectively referred to as “expressways”) pass through the regional mesh. When the area (regional mesh) is divided into 1 km (km) square, there are no interchanges such as expressways in many regional meshes even if the expressway passes through, and there are interchanges in specific areas To do. That is, in most areas, there are no vehicles entering and exiting the expressway and the area, and the areas where vehicles enter and exit are limited to areas where interchanges are provided. Also, areas with interchanges such as expressways are often different from traffic conditions compared to areas without interchanges. Can be evaluated more appropriately.
Furthermore, processing can be simplified by treating highways and the like as being not subject to analysis. As a result, it is possible to perform processing in which processing is unified between an area where there is an interchange and an area where there is no interchange, and it is easy to ensure continuity of analysis results.

The mesh tabulation QK will be described.
FIG. 8 is a diagram illustrating the aggregate QK for each area.
The graph shown in this figure is a graph showing the relationship between the number of traveling vehicles in the area and the aggregate traffic volume of those traveling vehicles.
In this graph, the vertical axis indicates the total traffic volume (vehicle km / hour), and the horizontal axis indicates the number of all traveling vehicles (vehicles) in the area. The total traffic volume shown on the vertical axis indicates the total traffic volume of all roads in the area.
The two types of scatter diagrams shown in this graph are classified into a weekday and a holiday (including Saturday. Hereinafter, unless otherwise specified, they are collectively referred to as “holiday”). Each plot shows the result totaled by time zone. In the example shown in this figure, the results are tabulated every hour.
On weekdays and holidays, there is a tendency that the distribution of the number of vehicles and the distribution of total traffic are different. It can be reproduced in this figure that the traffic trends on actual weekdays and holidays show different trends.

The arithmetic processing that led to this figure will be described.
The total traffic volume Q is obtained by totaling the products of (the length of each link j) and (the number of probe passages n _jτ ). The total traffic volume Q is shown as an equation (1).

In Expression (1), ΔT _i indicates the i-th time zone (1 hour width), τ indicates the total time (plus 15 minutes) belonging to the i-th time zone, and j indicates a set of links belonging to the mesh. , L _j indicates the length (km) of the link j, and n _jτ indicates the number of probes passing through the link j at the time τ.
Further, the area presence number K is obtained by totaling the products of (average travel time T _jτ at time τ of link j) and (number of probe passages n _jτ ). The area presence number K is expressed as Expression (2).

In Expression (2), T _jτ represents the average travel time (hour) at the time τ of the link j.
FIG. 9 is a diagram showing an aggregate QK approximate curve for each area from the scatter diagram shown in FIG.
The graph shown in this figure is a graph showing the relationship between the number of traveling vehicles in the area and the aggregate traffic volume of those traveling vehicles.
In this graph, the vertical axis indicates the total traffic volume (vehicle km / hour), and the horizontal axis indicates the number of all traveling vehicles (vehicles) in the area. The total traffic volume shown on the vertical axis indicates the total traffic volume of all roads in the area.
The approximate curve G31 is derived from a scatter plot of weekday plots, and the approximate curve G32 is derived from a scatter plot of holiday plots.
The approximate curves G31 and G32 are examples in which approximation is performed by a quadratic curve that protrudes upward along the origin based on the distribution state of each plot.
As shown in this figure, the approximate curves G31 and G32 showing weekdays and holidays are shown as approximate curves showing clearly different tendencies.

The calculation of the approximate curve will be described below.
A quadratic curve passing through the origin is expressed as equation (3), where the total traffic volume is defined as y and the number of areas present is defined as x.

In Equation (3), the coefficient a is a <0 because the graph is a convex curve. The coefficient b is b> 0 because the number of units is represented by a natural number.
When this approximated curve is evaluated using the least square method, an evaluation expression shown as Expression (4) can be defined, and a condition that minimizes the value is calculated.

  In order to derive the condition, f (a, b) is partially differentiated with respect to the variables a and b, and the condition in which the result is 0 (zero) is expressed as Expression (5).

  By arranging the equation (5) based on the equation (4), it is shown as the equation (6).

  By converting Equation (6) into a determinant, it is shown as Equation (7).

  In equation (7), the elements A, B and C of the matrix are shown as equation (8).

  Solving equation (7) leads to equation (9).

Subsequently, conversion to a parametric space and calculation of a congestion index will be described.
FIG. 10 is a diagram illustrating conversion from a quadratic function to a parametric space.
The graph shown in this figure is a graph showing a two-dimensional curve passing through the origin.
In this graph, the vertical axis represents y and the horizontal axis represents x. A quadratic curve passing through the origin shown in this graph is shown as Expression (10).

In Equation (10), the coefficient a is a <0 because the graph is a convex curve. The coefficient b is b> 0 because the number of units is represented by a natural number.
Expression (10) corresponds to the above-described expression (3) and represents a QK curve.
A QK curve expressed by a quadratic function is expressed as shown in Equation (11) through the parameter t.

In Expression (11), the parameter t is represented by 0 or more and 1 or less.
The QK approximate curve represented by the equation (11) is transformed to pass through each QK plot (x _i , y _i ) with the scale factor s _i and is represented as the equation (12).

  In Expression (12), based on the QK approximated curve that has been subjected to scaling adjustment, the parameter t is mediated and expressed as Expression (13).

In Expression (13), the parameter t is represented by 0 or more and 1 or less. Also, a parameter t _i corresponding to this plot is obtained. The parameter t _i , that is, the degree of moving away from the origin along the quadratic curve is defined as a congestion index.
Based on equation (13), the parameters t _i and s _i indicating the points on the curve scaled to pass through the point (x _i , y _i ) are shown as equation (14).

Based on the equation (14), the point (x * _i , y * _i ) indicated by the parameter t _i and the point (x _i , y _i ) on the QK curve before being transformed with the scale factor s _i Deriving the distance d _i is shown as equation (15).

In the equation (15), δs _i represents the Kronecker δ (delta) as shown in the equation (16).

In the equation (15), the distance d _i indicates a distance calculated by normalizing coordinates based on the vertical / horizontal size ratio (aspect ratio) of the quadratic curve.
Moreover, since the distance d _i indicates the degree of deviation from the average QK curve, it can be used as an index indicating unusualness.

Next, calculation of the congestion index will be described.
FIGS. 11A and 11B show the relationship between the congestion index and the deviation amount for weekdays and holidays, respectively.
In each graph, the vertical axis represents the deviation d from the QK curve, and the horizontal axis represents the congestion index t.
The plot of the scatter diagram shown in this graph is shown based on the tabulated results tabulated in units of 4 hours based on the total traffic volume.
As the value on the vertical axis moves away from “0”, the deviation from the approximate curve increases.
In the weekday distribution shown in FIG. 11A, there is little variation in the distribution from nighttime (22:00) to early morning (6 o'clock), and other time distributions have different divergence amounts depending on the day. It shows a similar trend.
On the other hand, in the holiday distribution shown in FIG. 11B, the distribution from nighttime (22:00) to early morning (6 o'clock) is small as in the weekday, but other time distributions tend to vary. Is shown.

The calculation of the two-dimensional normal distribution for each time zone from which this scatter diagram is derived will be described.
The parameters (t _i , d _i ) obtained for each QK plot i are plotted on the td plane for each time zone, and the parameters (μ _t , μ _d ,) of the two-dimensional normal distribution probability density function P (t, d) are plotted. σ _t ² , σ _d ² , ρ) are obtained.
An expression for deriving the event occurrence probability distribution P is shown in Expression (17).

In Expression (17), μ _t , μ _d , σ _t ² , and σ _d ² are the average and variance of parameters t and d, and are as shown in Expression (18).

  Moreover, in Formula (17), (rho) is a correlation coefficient and is shown as Formula (19).

The two-dimensional normal distribution probability density function P (t, d) shown in Expression (17) is applied to a minute region [t _j , ± Δt, d _j ± Δd] around the state (t _j , d _j ) of the day. The event occurrence probability p (t _j , d _j ) is obtained by integration, and is shown in Equation (20).

Finally, the entropy information amount I _j is obtained from the integration result of Expression (20).
An equation for deriving the entropy information amount I _J is Equation (21).

Subsequently, a procedure of traffic situation analysis processing performed by the traffic situation analysis apparatus 100 will be described.
FIG. 12 is a flowchart showing the procedure of the initialization process in the traffic condition analysis process.
The count target area information registration unit 111 records the input area information for specifying the count target area in the area information storage unit 112 by associating it with a regional information mesh obtained by dividing a digital map area in advance. When the recording of area information corresponding to all target areas is completed, the initialization process is terminated (step S11).

FIG. 13 is a flowchart showing the procedure of the analysis process in the traffic condition analysis process.
The information acquisition unit 121 acquires the number of link passages and link required time information at the time t in the target area. That is, the information acquisition unit 121 acquires the traffic situation data provided from the traffic information system 200 at a timing (time t) according to a predetermined cycle. The information acquisition unit 121 also includes map information of the target range stored in the area information storage unit 112, route information, area information that defines a target area indicating a unit for performing traffic condition analysis, and correspondence between the target area and the route. The information related to the range to be analyzed is acquired by referring to the information indicating the attachment. And the information acquisition part 121 acquires the number of vehicles which pass a target area, and the required time information concerning passage based on said information (step S21).

Area total QK state value calculation unit 122 calculates an area total QK state value at time t. That is, based on the information acquired by the information acquisition unit 121, the area total QK state value calculation unit 122 calculates an area total QK state value indicating the traffic situation of the area for each target area. The area total QK state value is calculated according to an arithmetic expression shown in Expression (1) to Expression (2).
The area total QK state value calculation unit 122 records the area total QK state value calculated for each target area in the area total QK state value storage unit 131. Recording in the area total QK state value storage unit 131 is performed according to the calculation of the area total QK state value, and time information corresponding to the information acquired by the information acquisition unit 121 is added and recorded in time series. (Step S22).

  The divergence amount calculation unit 123 performs projective transformation of the area total QK state value at time t along the total QK approximate curve, and calculates the divergence amount from the area congestion index and the approximate curve. That is, the divergence amount calculation unit 123 performs projective transformation on the area total QK state value calculated by the area total QK state value calculation unit 122 with respect to the total QK approximate curve derived based on the area total QK state value. Further, the divergence amount calculation unit 123 calculates the divergence amount from the area congestion index and the approximate curve from the result of the projective transformation (step S23).

  The change index calculation unit 124 calculates a change index (entropy information amount) from the two-dimensional normal distribution and the obtained area congestion index and divergence amount. That is, the change index calculation unit 124 changes from the two-dimensional normal distribution derived by the two-dimensional normal distribution parameter conversion unit 133, the area congestion index calculated by the deviation amount calculation unit 123, and the deviation amount from the approximate curve. A change index indicating the amount of entropy information indicating the degree of is calculated (step S24).

The congestion index output unit 125 outputs the calculated congestion index and change index for each area in accordance with the output form set stepwise (step S25).
The information acquisition unit 121 repeats the processing from step S21 when detecting that the predetermined time has been reached by an instruction from a timing unit provided in the apparatus. If the end of the analysis process is instructed, the process ends (step S26).

FIG. 14 is a flowchart illustrating the procedure of the approximate curve calculation process in the traffic condition analysis process.
The deviation amount calculation unit 123 records the calculated area aggregation QK state value in the area aggregation QK state value storage unit 131, and performs an update process of the stored data. This update process is performed according to the data acquisition in step S21, and the calculation result of step S23 is reflected (step S31).

  The total QK approximate curve unit 132 refers to the area total QK state value for each area stored by the area total QK state value storage unit 131 and calculates a total QK approximate curve for each area. This approximate curve calculation process is repeatedly performed with a relatively long cycle. For example, the repetition cycle is 1 week. The calculated result is recorded and updated as data indicating a new aggregate QK approximate curve (step S32).

  The two-dimensional normal distribution parameter conversion unit 133 performs projective transformation of the area total QK state value stored in the area total QK state value storage unit 131 along the total QK approximate curve for each time zone, and generates a two-dimensional normal distribution. Calculate the parameters. The calculation of the parameters of the two-dimensional normal distribution is repeatedly performed with a relatively long cycle. For example, it is performed in accordance with step S32, and the repetition cycle is one week. The calculated result is recorded and updated as data indicating a new two-dimensional normal distribution parameter (step S33).

In the embodiment of the present invention, the total QK approximate curve portion 132 is an area traffic indicating the number of vehicles existing in the area and the number of traveling vehicles passing through the area for each predetermined area unit. An approximate curve (aggregate QK approximate curve) showing the relation of the state information as a function that is convex upward is calculated. The deviation amount calculation unit 123 calculates a congestion index indicating the degree of difficulty in passing the area based on the positional relationship with the aggregate QK approximate curve.
Thereby, the traffic condition analysis apparatus 100 can provide the traffic condition for each area without being limited to information on the main trunk road and the like.

The change index calculation unit 124 calculates a change index from the difference between the aggregate QK approximate curve indicated by the upward convex function as a reference and the area traffic state information.
Thereby, the traffic condition analysis apparatus 100 can detect the difference from the normal time by referring to the calculated change index.

The aggregate QK approximate curve unit 132 calculates an approximate curve indicated by a function that is convex upward as a quadratic curve.
Thereby, the traffic condition analysis apparatus 100 can detect the difference from the normal time by referring to the calculated change index while reducing the processing load by treating the approximate curve as a quadratic curve.

The total QK approximate curve unit 132 sets the area traffic state information as the total traffic density indicated by the total travel time of the vehicle traveling per unit time on the road selected as the target road in the area.
Thereby, the total QK approximation curve part 132 can derive | lead-out area traffic state information from the total traffic density shown by the total travel time of the vehicle which drive | works the road of an object area per unit time.

The total QK approximate curve unit 132 sets the area traffic state information as the total traffic amount indicated by the total travel distance of the vehicle traveling per unit time on the road selected as the target road in the area.
Thereby, the total QK approximation curve part 132 can guide | lead area traffic state information from the total traffic shown by the total travel distance of the vehicle which drive | works the road of an object area per unit time.

The congestion index output unit 125 outputs a change in the congestion index or change index for each area in a stepwise index.
Thereby, by displaying the congestion index or the change index stepwise, it is possible to easily distinguish the displayed contents.

The congestion index output unit 125 maps an area on a map and displays the area according to a difference in color or density of the area based on a stepwise index derived from the congestion index or change index derived for each area.
Thereby, the information of the congestion index or the change index shown on the map is shown by the difference in the color or density of the area which is displayed stepwise, so that the displayed content can be easily discriminated.

The present invention is not limited to the above embodiments, and can be modified without departing from the spirit of the present invention.
The change index calculation unit 124 has shown that the upward convex function as a reference is approximated by a quadratic curve. For example, only the function indicated by the polygonal line indicating the upward convex shape or the congestion area A function that approximates with a monotonously decreasing function that protrudes downward may be used. The congestion area is an area on the right side of the point indicating the maximum value of the upward convex function, that is, an area on the right side of the mountain.
In addition, although the characteristics of changes in traffic volume compiled for each area are shown separately for weekdays and holidays, they can be classified according to day of the week or period as necessary. Moreover, it can classify | categorize by arbitrary time also about the classification | category for every time slot | zone.
In addition, the area as the aggregation unit is merely an example of division according to the digital map as a regional mesh, and any shape and size can be selected, and each area has the same shape or the same size. You are not limited to doing that.

  In addition, the above-mentioned traffic condition analysis apparatus 100 has a computer system inside. The processing procedure for providing the traffic situation described above is stored in a computer-readable recording medium in the form of a program, and the above processing is performed by the computer reading and executing this program. Here, the computer-readable recording medium means a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, a semiconductor memory, or the like. Alternatively, the computer program may be distributed to the computer via a communication line, and the computer that has received the distribution may execute the program.

DESCRIPTION OF SYMBOLS 100 Traffic condition analyzer 111 Total object area information acquisition part 112 Area information storage part 121 Information acquisition part 122 Area total QK state value calculation part 123 Deviation amount calculation part 124 Change index calculation part 125 Congestion index output part 131 Area total QK state value Storage unit 132 Aggregated QK approximate curve unit 133 Two-dimensional normal distribution parameter conversion unit

Claims (8)

For each predetermined area unit, an approximate curve indicating the relationship between the number of vehicles existing in the area and area traffic state information indicating the number of traveling vehicles of vehicles passing through the area as a function that is convex upward An approximate curve calculation unit for calculating
A congestion index calculation unit that calculates a congestion index indicating the degree of difficulty in passing the area based on the positional relationship with the approximate curve;
A change index calculating unit that calculates a change index from a difference between the approximate curve indicated by the upward convex function as a reference and the area traffic state information;
A traffic condition analyzing apparatus comprising: The approximate curve calculation unit includes:
The traffic condition analysis apparatus according to claim 1, wherein projective transformation is performed along the approximate curve indicated by the upward convex function to calculate a parameter of a two-dimensional normal distribution. The approximate curve calculation unit includes:
The area traffic state information is a total traffic density indicated by a total travel time of a vehicle traveling per unit time on a road selected as a target road in the area. Traffic condition analysis device. The approximate curve calculation unit includes:
The area traffic state information is a total traffic amount indicated by a total travel distance of a vehicle traveling per unit time on a road selected as a target road in the area. Traffic condition analysis device. A congestion index output unit that outputs the change of the congestion index or the change index for each area in a stepwise index;
The traffic condition analysis apparatus according to any one of claims 1 to 4, further comprising: The congestion index output unit is:
The area is mapped on a map and displayed by a difference in color or density of the area based on a congestion index derived for each area or a stepwise index derived from the change index. Item 6. The traffic condition analysis device according to Item 5. For each predetermined area unit, an approximate curve indicating the relationship between the number of vehicles existing in the area and area traffic state information indicating the number of traveling vehicles of vehicles passing through the area as a function that is convex upward The procedure to approximate with
A procedure for calculating a congestion index indicating a degree of difficulty in passing the area based on a positional relationship with the approximate curve;
A procedure for calculating a change index from a difference between the approximate curve indicated by the upward convex function as a reference and the area traffic state information;
A traffic condition analysis program characterized by causing a computer to execute. For each predetermined area unit , the approximate curve calculation unit has an upwardly convex relationship between the number of vehicles existing in the area and the area traffic state information indicating the number of vehicles traveling through the area. A step of approximating with an approximate curve indicated by a function;
A congestion index calculating unit calculating a congestion index indicating a degree of difficulty in passing the area based on a positional relationship with the approximate curve;
A change index calculating unit calculating a change index from a difference between the approximate curve indicated by the upward convex function as a reference and the area traffic state information;
The traffic condition analysis method characterized by having.