JP2009019920A - Route search device, traffic simulation apparatus, pedestrian behavior prediction system, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To search for a route which a pedestrian actually selects. <P>SOLUTION: A route candidate creation means creates a plurality of route candidates from the place of departure up to a destination of the pedestrian based on map information. For each route candidate created by the route creation means, a calculation means calculates an overall degree of danger according to the kinds of places to move to existing on the route candidate, based on a degree of danger to be assumed on the occasion of movement predetermined for each of a plurality of kinds of places for the pedestrian to move to. Accordingly, the route which the pedestrian actually selects can be searched for, since the route selection is performed considering the individual degrees of danger of the route candidates. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a route search device, a traffic simulation device, a pedestrian behavior prediction device, and a program, and more particularly to a route search device that searches for a route to a pedestrian or bicycle destination, and a simulation of the behavior of the pedestrian or bicycle. The present invention relates to a traffic simulation apparatus, a pedestrian behavior prediction apparatus that predicts the behavior of a pedestrian or a bicycle, and a program.

  2. Description of the Related Art Conventionally, methods for determining a walking route of a pedestrian based on the pedestrian's personal attributes and walking conditions are known (for example, Patent Documents 1 and 2).

There is also known an in-vehicle device that estimates the pedestrian's behavior from the current position and past trajectory of the pedestrian and determines the risk of a pedestrian accident (Patent Document 3).
Japanese Patent No. 3370555 JP-A-8-202982 JP 2006-330822 A

  However, the techniques described in Patent Documents 1 and 2 do not take into account risks such as the risk of encountering an accident, and thus there is a problem that a route that a pedestrian can actually select cannot be determined.

  In addition, in the technique described in Patent Document 3, a judgment from the pedestrian side, for example, a judgment that considers a risk such as “Let's cross the crosswalk in another direction because the vehicle passes the crosswalk in front” is not performed. Therefore, there is a problem that the route that the pedestrian actually selects cannot be estimated.

  The present invention has been made to solve the above problems, and is a route search device, a traffic simulation device, a pedestrian behavior prediction device, and a program capable of searching for a route that a pedestrian or bicycle actually selects. The purpose is to provide.

  In order to achieve the above object, the route search device according to the first invention comprises a route candidate generating means for generating a plurality of route candidates from a pedestrian or bicycle departure point to a destination based on map information; On each of the route candidates generated by the route generation means on the basis of the risk level when moving in advance for each of a plurality of types of pedestrians or bicycles, on the route candidates. Calculating means for calculating the total risk according to the type of the moving location existing in the vehicle, and output means for outputting the total risk of each of the route candidates calculated by the calculation means. Yes.

  According to a second aspect of the invention, there is provided a program for generating route candidates from a starting point of a pedestrian or a bicycle to a destination based on map information. For each of the route candidates generated by the route generation means based on the predetermined risk when moving for each of a plurality of types, the type of the moving location existing on the route candidate A program for functioning as a calculation means for calculating the total risk according to the output, and an output means for outputting the total risk of each of the route candidates calculated by the calculation means.

  According to the first invention and the second invention, the route candidate generation means generates a plurality of route candidates from the starting point of the pedestrian or the bicycle to the destination based on the map information. And, for each of the route candidates generated by the route generation means, based on the risk when moving in advance for each of a plurality of types of pedestrian or bicycle movement locations by the calculation means, The total risk according to the type of moving location present on the route candidate is calculated, and the total risk of each of the route candidates calculated by the calculation means is output by the output means.

  Therefore, for each of the route candidates, the route is selected in consideration of the risk of each of the route candidates by calculating the total risk according to the type of moving location existing on the route candidate. Therefore, it is possible to search for a route actually selected by the pedestrian or the bicycle.

  The output means according to the first invention can display a plurality of route candidates generated by the route generation means and the total risk of each of the route candidates. Thereby, the operator can select a route in consideration of the displayed total risk.

  The route search device according to the first invention is based on the total risk of each of the route candidates output by the output unit, and the route of the pedestrian or the bicycle from the plurality of route candidates generated by the route generation unit. Further, a route selection means for selecting can be included. Thereby, it is possible to search for a route that the pedestrian or the bicycle actually selects.

  The types of pedestrian or bicycle travel locations described above can include pedestrian crossings, walkway-separated walkways, non-pedestrian walkways, and roadways.

  Further, the degree of risk for the roadway can be determined for each of the vehicle traffic per unit time of the roadway or each of a plurality of average vehicle speeds of the vehicle traveling on the roadway. Accordingly, it is possible to calculate the degree of danger of the roadway where the pedestrian or the bicycle moves, considering the vehicle traveling on the roadway.

  The route search device described above is based on at least one of the position of the vehicle with respect to the current position of the pedestrian or the bicycle and the speed of the vehicle, for each of the route candidates generated by the route generation means. Based on the collision risk calculation means for calculating the collision risk that the vehicle collides with the vehicle, the total risk of each of the route candidates output by the output means, and the collision risk calculated by the collision risk calculation means In addition, route selection means for selecting a pedestrian or bicycle route from a plurality of route candidates generated by the route generation means can be further included. Accordingly, the route of the pedestrian or the bicycle can be selected in consideration of the collision risk with the vehicle that runs near the current position of the pedestrian or the bicycle.

  The route candidate generation means generates and generates a plurality of route candidate points that pass through the route from the departure point to the destination and a plurality of links that connect two route candidate points among the plurality of route candidate points. The plurality of links are used to generate a plurality of route candidates, and the calculation means calculates and calculates a risk level for each of the generated plurality of links according to the type of moving location existing on the link. Based on the respective risk levels of the plurality of links, the total risk level can be calculated for each of the route candidates generated by the route generation means according to the type of moving location present on the route candidate. .

  The route search apparatus calculates a travel time for each of the route candidates generated by the route candidate generation means, and moves in advance according to each of a plurality of types of pedestrians or bicycles. For each of a plurality of route candidates, a total load amount corresponding to the type of moving location present on the route candidate is calculated based on the load amount, and based on the calculated travel time and total load amount For each of a plurality of route candidates, a cost calculation unit that calculates a cost when moving, a risk level output by the output unit, and a cost calculated by the cost calculation unit are generated by the route generation unit. Route selection means for selecting a route of a pedestrian or a bicycle from a plurality of route candidates. Accordingly, the route can be selected in consideration of the risk and cost of each of the route candidates, so that the route actually selected by the pedestrian or the bicycle can be searched.

  Further, the total load amount can be calculated based on the load amount, the weather, whether or not the vehicle moves while running, and whether or not the movement is in violation of the law. Accordingly, the route can be selected in consideration of the load caused by the weather, the load caused by running, and the load caused by violating the law.

  In addition, the route selection unit is configured to determine whether the pedestrian or the bicycle is based on the total risk level output by the output unit, the cost calculated by the cost calculation unit, and the personal characteristics of the set pedestrian or bicycle. A route can be selected. As a result, the route can be selected in consideration of the risk, cost and personal characteristics of the pedestrian or bicycle, so that the route actually selected by the pedestrian or bicycle can be searched.

  According to a third aspect of the present invention, there is provided a route search device for generating a plurality of waypoints on a route candidate from a current position of a pedestrian or a bicycle to a destination based on map information, and a pedestrian Or, for each of the waypoints generated by the waypoint generation means based on the predetermined risk when moving for each of a plurality of types of bicycle moving places, the waypoint from the current position Calculation means for calculating the total risk according to the type of the moving location existing on the route up to, and output means for outputting the total risk of each of the waypoints calculated by the calculation means It is configured.

  According to the route search device according to the third aspect of the present invention, the waypoint generating means generates a plurality of waypoints existing on the route candidates from the current position of the pedestrian or the bicycle to the destination based on the map information. . And, for each of the waypoints generated by the waypoint generation means, based on the risk when moving in advance for each of a plurality of types of pedestrian or bicycle movement place by the calculation means, The total risk according to the type of moving location existing on the route from the current position to the waypoint is calculated, and the total risk of each waypoint calculated by the calculation means is output by the output means.

  Therefore, for each waypoint, calculate the total risk according to the type of travel location that exists on the route to the waypoint, and select the route in consideration of the risk of each waypoint. Therefore, it is possible to search for a route actually selected by the pedestrian or the bicycle.

  According to a fourth aspect of the present invention, a traffic simulation apparatus travels a route search device including the route selection means and a vehicle model that models a vehicle, and models a pedestrian according to the route selected by the route search device. And a simulation means for simulating a traffic state by moving a modeled pedestrian model or a modeled bicycle model.

  According to the traffic simulation device of the fourth invention, the route search device selects a route for a pedestrian or bicycle from a plurality of route candidates. Then, the vehicle model that models the vehicle or the bicycle model that models the bicycle is run by the simulation means, and the pedestrian model or bicycle that models the pedestrian is modeled according to the route selected by the route search device. The bicycle model is moved to simulate traffic conditions.

  Therefore, for each of the route candidates, the total risk according to the type of moving location present on the route candidate is calculated, and the pedestrian model is determined according to the route selected in consideration of the risk of each of the route candidates. Alternatively, by moving the bicycle model, it is possible to simulate a traffic state in which the pedestrian model or the bicycle model is moved according to a route actually selected by the pedestrian or the bicycle.

  The traffic simulation device according to the fourth invention stores driver accident information representing a driver or vehicle event that causes an accident for each driver attribute and accident environment, and also includes a pedestrian attribute or bicycle attribute, and each accident environment. Storage means for storing pedestrian accident information representing an event of a pedestrian or a bicycle that causes an accident, and a driver or vehicle event that is an accident induction factor based on the driver accident information stored in the storage means A driver event generating means for generating a pedestrian event, and a pedestrian event generating means for generating a pedestrian or bicycle event that is an accident inducing factor based on pedestrian accident information stored in the storage means, The means includes an accident inducing factor based on the driver accident information stored in the storage means and the road environment in which the vehicle model travels. When generating an event of a driver or a vehicle, vehicle behavior determination means for determining the behavior of a vehicle model so as to generate the event, pedestrian accident information stored in the storage means, and the route search In the case of generating a pedestrian or bicycle event that causes an accident based on the route selected by the device and the road environment where the pedestrian model walks or the road environment where the bicycle model runs, the event is Pedestrian behavior determining means for determining the behavior of a pedestrian model or a bicycle model so as to be generated, running the vehicle model based on the behavior of the vehicle model determined by the vehicle behavior determining means, and The pedestrian model or bicycle model is moved based on the behavior of the pedestrian model or bicycle model determined by the pedestrian behavior determining means. Te, it is possible to simulate the traffic state. As a result, the occurrence of an accident can be simulated.

  Further, the traffic simulation device described above compares the result of simulating the traffic state by mounting a safety system on the vehicle model and the result of simulating the traffic state without mounting the safety system on the vehicle model, The vehicle behavior determination means further includes evaluation means for evaluating the effect of the safety system, and the vehicle behavior determination means is based on driver accident information stored in the storage means, a road environment in which the vehicle model travels, and whether or not the vehicle model safety system is installed. Thus, the behavior of the vehicle model can be determined. Thereby, the effect of the safety system mounted on the vehicle can be evaluated.

  In addition, the above traffic simulation device is a result of simulating a traffic state by carrying a safety system in a pedestrian model or a bicycle model, and a result of simulating a traffic state without carrying a safety system in a pedestrian model or a bicycle model. The pedestrian behavior determining means further includes pedestrian accident information stored in the storage means, and the route selected by the route search device. The behavior of the pedestrian model or the bicycle model can be determined based on the road environment where the pedestrian model walks or the road environment where the bicycle model travels, and the presence or absence of the pedestrian model or the bicycle model safety system. Thereby, the effect of the safety system carried by a pedestrian or a bicycle can be evaluated.

  A pedestrian behavior prediction apparatus according to a fifth invention predicts the behavior of a prediction target pedestrian or bicycle based on the route search device including the route selection unit and the route selected by the route selection unit. And prediction means.

  The pedestrian behavior prediction apparatus according to the fifth aspect of the present invention selects a pedestrian or bicycle route from a plurality of route candidates by a route search device. Then, the behavior of the pedestrian or bicycle to be predicted is predicted by the prediction means.

  Therefore, for each of the route candidates, calculate the total risk according to the type of moving location existing on the route candidate, select the route in consideration of the risk of each of the route candidates, and walk By predicting the behavior of a person or a bicycle, the behavior of a pedestrian or a bicycle can be predicted with high accuracy.

  According to a sixth aspect of the present invention, there is provided a program for generating route candidate generation means for generating a plurality of route candidates from a starting point of a pedestrian or bicycle to a destination based on map information, For each of the route candidates generated by the route generation means based on the predetermined risk when moving for each of a plurality of types, the type of the moving location existing on the route candidate Calculating means for calculating the total risk according to the output means, output means for outputting the total risk of each of the route candidates calculated by the calculating means, and total of each of the route candidates output by the output means Based on the degree of risk, the route selection means for selecting the route of the pedestrian or the bicycle from the plurality of route candidates generated by the route generation means, and the vehicle are modeled. Both models are run, and a pedestrian model that models a pedestrian or a bicycle model that models a bicycle is moved according to the route selected by the route selection unit, and functions as a simulation unit that simulates a traffic state. It is a program for.

  As described above, according to the route search device and program of the present invention, for each route candidate, by calculating the total risk according to the type of moving location present on the route candidate, or For each waypoint, consider the risk of each of the route candidates or the risk of each waypoint by calculating the total risk according to the type of travel location on the route to the waypoint. In addition, since the route can be selected, an effect that the route that the pedestrian or the bicycle actually selects can be searched.

  In addition, according to the traffic simulation device and the program according to the present invention, for each route candidate, the total risk according to the type of moving location existing on the route candidate is calculated, and the risk of each route candidate is calculated. By moving the pedestrian model or the bicycle model according to the route selected in consideration of the degree, it is possible to simulate the traffic state in which the pedestrian model or the bicycle model is moved according to the route actually selected by the pedestrian or the bicycle. The effect that it can be obtained.

  In addition, according to the pedestrian behavior prediction device according to the present invention, for each of the route candidates, the total risk according to the type of the moving location existing on the route candidate is calculated, and each of the route candidates is calculated. By selecting a route in consideration of the degree of risk and predicting the behavior of the pedestrian or the bicycle, it is possible to predict the behavior of the pedestrian or the bicycle with high accuracy.

  Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings. In the first embodiment, a case where the present invention is applied to a pedestrian simulation apparatus that virtually walks a pedestrian model that models a pedestrian on a road will be described.

  As shown in FIG. 1, the pedestrian simulation apparatus 10 according to the first embodiment includes a walking environment database 12, a pedestrian database 14, and a pedestrian simulation unit 16. The pedestrian simulation device 10 is realized by, for example, a computer connected to a network, and a computer stores a program for executing a simulation processing routine described later.

  The walking environment database 12 stores walking environment information indicating road maps, signals, buildings, weather, light and darkness during walking, and the like.

  The pedestrian database 14 stores the pedestrian attributes of each pedestrian. Specifically, the pedestrian database 14 stores a departure place, a destination, a departure time, age, sex, physical ability, consciousness level, and the like for each pedestrian.

  The pedestrian simulation unit 16 reproduces the behavior of each pedestrian based on the data in the walking environment database 12 and the pedestrian database 14. The pedestrian simulation unit 16 includes a walking environment creation update unit 18, a route candidate point link generation unit 20, and a pedestrian behavior calculation unit 22.

  The walking environment creation update unit 18 creates a road model based on walking environment information indicating a road map stored in the walking environment database 12, and creates a walking environment model based on other walking environment information. In addition, a pedestrian model is arranged on the road model based on the data from the pedestrian database 14. Further, the arrangement position of the pedestrian model is updated based on the behavior of the pedestrian model calculated by the pedestrian behavior calculation unit 22.

  The route candidate point link generation unit 20 generates a pedestrian model on the road model based on the data such as the departure place and the destination from the pedestrian database 14 and the road model and the walking environment model from the walking environment creation update unit 18. Generates a transit candidate point and a link.

  As shown in FIG. 2 (A), on the basis of the pedestrian's starting point O and destination D, the pedestrian model is located at a point on the road model where the facility entrance, the intersection angle, and the boundary of the roadway and sidewalk attributes are different. Create multiple route candidate points through which. If the distance between route candidate points is long, route candidate points are also created between them.

  Further, as shown in FIG. 2B, a plurality of links are created by connecting two candidate via points. Even in a direction away from the departure point O to the destination D, for example, no link is generated between candidate via points separated by 200 m or more.

  In addition, when a link is generated at a signalized intersection as shown in FIG. 3A, two via candidate points are connected so as to pass on a pedestrian crossing or in the intersection as shown in FIG. 3B. To create a link.

  As shown in FIG. 4, when there is an obstacle (for example, a guardrail, a separation zone, a roadside tree) between the route candidate points, on the extension line connecting the route candidate point that is the start point of the link and the end point of the obstacle. A route candidate point is further created, and a link is created using the created route candidate point.

  The pedestrian behavior calculation unit 22 determines an optimum route using the route candidate points and links generated by the route candidate point link generation unit 20. When the route is determined, the surrounding environment such as a signal and the vehicle are recognized. Based on their existence (considering the speed in the case of moving objects) and the distance to each, the behavior such as the speed and the optimum route are determined. To be judged. Further, based on this determination, the behavior of the pedestrian is calculated, and the pedestrian position is updated by the walking environment creation update unit 18 based on the calculated behavior.

  When determining the route, the optimal route from the starting point to the destination is searched using the Dikstra method. If the distance to the destination is long, an optimum route to a waypoint on the way may be searched.

In determining the route, an optimum route is selected from a plurality of route candidates based on the utility U _kj of the pedestrian j for each of the route candidates to the destination. The utility U _kj is calculated by the following equation (1) using the cost C _ij when each link i existing on the route candidate moves and the risk R _ij indicating the danger level when moving.

Here, k represents a set of links existing on the route candidate.

  As in the above equation (1), the utility of each of the route candidates is obtained by the sum of the total cost and the total risk of each link existing on the route candidate.

Here, the cost C _ij is calculated by the following equation (2) using the travel time T _ij and the load B _ij .

  As in the above equation (2), the total cost of each route candidate is obtained by the sum of the total travel time of each link existing on the route candidate and the total load.

In addition, even if the route is the same, if there are options for the movement method (whether walking or running), the route is different from each other. For example, if the route candidate is a route from 30m before the pedestrian crossing or the roadway scheduled to cross, and a link is generated at a signalized intersection as shown in FIG. 5, for example, FIG. As shown, each of link I ₆ and link I ₈ is not a single link, but for I ₆ there are two links (I _6Run , I _6Walk ) and for I ₈ there are two links (I _8Run , I _{8 Walk} ).

Further, the travel time T _ij is calculated by a function of the link length and the moving speed, as shown in the following equation (3).
Travel time T _ij = f (link length, moving speed) (3)
However, the moving speed is determined based on pedestrian attributes such as age, physical ability, and consciousness level. Also, the estimated time for waiting for a signal or waiting for passing a vehicle is added to the travel time.

Here, a method for calculating travel time will be described. First, depending on the age of the pedestrian attribute of the pedestrian, the walking speed V _Walk or the running speed V _Run is determined from the predetermined relationship between age and speed as shown in FIG. Then, considering the physical ability and consciousness level of the pedestrian attribute, the walking speed or the traveling speed as the moving speed is calculated by the following formulas (4) and (5).
V _ij = V _Walk + α × (consciousness level) + β × (physical ability level) (4)
V _ij = V _Run + α × (consciousness level) + β × (physical ability level) (5)

  However, α and β are predetermined parameters, and the consciousness level is determined from a predetermined relationship as shown in FIG. 8 according to the consciousness level of the pedestrian attribute. It is determined from a predetermined relationship as shown in FIG. 9 according to the physical ability of the pedestrian attribute.

Next, using the following equation (6), the travel time is calculated by adding the estimated times of waiting for traffic lights and waiting for passing vehicles to the travel time based on the moving speed V _ij and the link length L _i .
T _ij = L _i / V _ij
+ Γ _signal × (prediction time until blue _signal × prediction error)
+ Γ _car × (predicted time until vehicle passes × predictive error) (6)

However, γ _signal is 1 when waiting for a signal, 0 when not waiting, and γ _car is 1 when waiting for a vehicle to pass, and 0 when not waiting. In addition, a prediction error is considered in each prediction time of waiting for a signal and waiting for passing a vehicle, and the prediction error is calculated using the following equation (7) in consideration of a consciousness level, age, and experience.
Prediction error = α x (consciousness level)
+ Β × (age level) + γ × (experience level) (7)
However, α, β, and γ are predetermined parameters, and the consciousness level is determined from a predetermined relationship as shown in FIG. 10 according to the consciousness level of the pedestrian attribute, and the age level is Depending on the age of the pedestrian attribute, it is determined from a predetermined relationship as shown in FIG. The experience level is determined from a predetermined relationship as shown in FIG. 12 according to the experience of the pedestrian attribute.

Next, the load B _ij is <1> walking environment such as the height of a road such as a pedestrian bridge or pavement, <2> weather, <3> whether or not it is a moving movement, and <4> whether the movement is a violation. It expresses the physical and mental load depending on whether or not, and is calculated as a function of walking environment, link length, moving speed, physical ability, age, and legal compliance tendency as shown in the following equation (8).
B _ij = f (walking environment, link length, movement speed, physical ability, age, legal compliance tendency)
= L _i / V _ij
× ((weight of walking environment load) + α × (physical ability level of walking environment load) + β × (age level of walking environment load) + a × η × (weight of weather load)
+ B × (γ × (Running load physical ability level) + δ × (Running load age level))
+ Κ x (weight of violating behavior load) x (legal compliance level))
... (8)

  However, a is 0.0 when the moving place is indoor, 0.5 when the moving place is outdoor and has a roof, and 1.0 when the moving place is outdoor and there is no roof. . b is 0 for a walking movement and 1 for a running movement. In addition, α, β, γ, δ, η, and κ are predetermined parameters.

  Further, as shown in FIG. 13, the weights of the walking environment load are as follows: types of moving places “paved, flat”, “slope (steep)”, “slope (normal)”, “slope (loose)”, “pedestrian bridge” ”,“ Gravel road ”,“ width-small sidewalk ”, and“ stairs ”. The physical ability level of the walking environment load is determined from a predetermined relationship as shown in FIG. 14 according to the physical ability of the pedestrian attribute, and the age level of the walking environment load is determined according to the age of the pedestrian attribute. , Determined from a predetermined relationship as shown in FIG.

  Further, as shown in FIG. 16, the weight of the weather load is as follows for each of the types of weather “rain”, “snow”, “mist”, “cloud”, “sunny”, “temperature”, and “humidity”. The weighting factor is predetermined according to the weather level (large, medium, small), and the physical ability level of the running load is predetermined as shown in FIG. 17 according to the physical ability of the pedestrian attribute. The age level of the running load determined from the relationship is determined from a predetermined relationship as shown in FIG. 18 according to the age of the pedestrian attribute.

  Further, as shown in FIG. 19, the weight of the violation behavior load is predetermined for each of the violation content types “red light”, “flashing signal”, “crossing the roadway”, and “crossing within the intersection”. It is a weighting factor. Further, the legal compliance level is determined from a predetermined relationship as shown in FIG. 20 according to the legal compliance tendency of the pedestrian attribute.

Next, a risk calculation method will be described. The risk R _ij for the link i of the pedestrian model j is calculated using a static risk function f _u that does not change, such as a dangerous walking environment, as in the following equation (9).
R _ij = _fu (walking environment, link length, moving speed, age, experience)
= Li / Vij
× (Walking environment weight + α × experience level + β × age level) (9)

  However, α and β are predetermined parameters, and the weight of the walking environment is, as shown in FIG. 21, the types of moving places “pedestrian crossing”, “roadway (small traffic)”, “roadway (traffic volume) It is a weighting factor determined in advance for each of “Large)”, “Waveway (within the intersection)”, “Walking sidewalk”, and “Walking sidewalk”. As shown in FIG. 21, the weighting factor of the type of “location” of the moving location is determined according to the traffic volume (small, large) of the vehicle traveling on the roadway. Note that the weighting factor for the type of moving place “roadway” may be determined according to the average speed (small, large) of the vehicle.

  The experience level is determined from a predetermined relationship as shown in FIG. 22 according to the experience of the pedestrian attribute, and the age level is determined in advance as shown in FIG. 23 according to the age of the pedestrian attribute. It is determined from the established relationship.

  Risks are also determined by further considering the pedestrian congestion rate, pedestrians present at the moving location, whether or not the moving location is separated from the bicycle road, the amount of bicycle traffic at the moving location, the rate of mixed large vehicles on the roadway, etc. May be calculated.

  From the above, the total risk for the route candidates is the total risk calculated by the above equation (9) according to the type of travel location corresponding to the link existing on the route candidates.

The pedestrian behavior calculation unit 22 calculates a cost C _ij and a risk R _ij for each of the links i related to the pedestrian j generated by the route candidate point link generation unit 20, and for each of the route candidates, The cost C of all the links corresponding to the candidates is summed, the risk R is summed, and the utility U is calculated by the sum of the sum of the costs C and the sum of the risks. Then, based on the utility U of each of the route candidates, a route candidate having the smallest utility U is searched, and the searched route candidate is determined as an optimum route.

  Next, the operation of the pedestrian simulation apparatus 10 according to the first embodiment will be described. First, when an operator inputs a simulation start time and an end time to the pedestrian simulation apparatus 10 and inputs a simulation start instruction, the pedestrian simulation apparatus 10 executes a simulation processing routine shown in FIG. .

  First, in step 100, a road model is generated based on walking environment information indicating a road map in the walking environment database 12, and a walking environment model at a simulation start time is generated based on walking environment information indicating signals, buildings, and the like. In step 102, based on the data in the pedestrian database 14, a pedestrian model whose start time corresponds to the departure time is arranged at the departure place on the road model, and in step 104, for each pedestrian model, A plurality of route candidate points from the departure point to the destination are generated on the road model, and a plurality of links are generated on the road model based on the generated route candidate points.

  In the next step 106, the road environment recognized by the pedestrian model is acquired for the target pedestrian model, and in step 108, determination processing is performed based on the recognized road environment acquired in step 106.

  In the determination process of step 108, the sum of risk and the sum of cost are calculated for each route candidate based on the recognized road environment, and the utility is calculated based on the sum of risk and the sum of cost. . Then, a route candidate having the minimum utility is searched, and the searched route candidate is determined as the optimum route. In step 110, the behavior of the target pedestrian model is calculated based on the determined optimum route.

  In the next step 112, it is determined whether or not the processing of step 106 to step 110 has been executed for all the pedestrian models arranged on the road model, and not executed for all the pedestrian models. In step 106, the process of recognition, determination, and behavior is executed for the next pedestrian model. On the other hand, when the processing of Step 106 to Step 110 is executed for all pedestrian models, the process proceeds to Step 114, and the current time is updated to the next time.

  In step 116, the walking road environment model is updated to the walking road environment model at the updated time, and the placement position of the pedestrian model is based on the behavior of the pedestrian model calculated in step 110. Update. Note that the pedestrian model that has arrived at the destination is extinguished at this point.

  In the next step 118, it is determined whether or not it is an end time. If the current time is the end time, the simulation processing routine is terminated. If the current time is not the end time, in step 120, If it is determined whether or not to add a pedestrian model and there is no pedestrian whose updated current time is the departure time based on the data in the pedestrian database 14, the process returns to step 106. If there is a pedestrian whose current time is the departure time, it is determined that a pedestrian model is to be added, and in step 122, the pedestrian model whose current time is the departure time is placed at the departure place on the road model.

  In step 124, for each of the pedestrian models newly arranged in step 122, a plurality of route candidate points and a plurality of links from the departure point to the destination are generated on the road model, and the flow goes to step 106. Return.

  As described above, when the simulation processing routine is executed, for each pedestrian model of the pedestrian indicated by the data in the pedestrian database 14, each pedestrian model is processed based on the processes of recognition, judgment, and behavior. A simulation is performed so that the optimum route is selected and each pedestrian model is moved toward the destination according to the selected route.

  As described above, according to the pedestrian simulation apparatus according to the first embodiment, for each of the route candidates, the sum of the risks in consideration of the types of moving locations corresponding to the links existing on the route candidates. , And moving the pedestrian model according to the route actually selected by the pedestrian by moving the pedestrian model according to the route selected in consideration of the total risk for each of the route candidates. It can be simulated.

  In addition, since the route can be selected in consideration of the total risk and the total cost of the route candidates, the traffic state in which the pedestrian model is moved according to the route actually selected by the pedestrian is simulated. can do.

  Moreover, pedestrian behavior close to reality can be simulated by simulating the route and behavior of the pedestrian in consideration of the personal attributes of the pedestrian model, the ability, and the risk based on the surrounding road environment.

  In addition to the preference of the route and behavior according to personal attributes, the optimal route is judged and the behavior is calculated considering the physical load of the environment such as slopes, so the pedestrian route and behavior closer to reality Can be reproduced.

In the above embodiment, the case where the utility is calculated by the sum of the sum of the costs and the sum of the risks has been described as an example. However, the utility is obtained by weighting each of the costs and the risks. You may make it calculate. In this case, a weighting factor is set according to the personal characteristics of the pedestrian, for example, the age, and the route candidate (k is a link existing on the route candidate) of the pedestrian model j by the following equation (10). The utility U _kj may be calculated.

However, w _Cj is a weighting coefficient related to the cost according to the age of the pedestrian model j, and _wRj is a weighting coefficient related to the risk according to the age of the pedestrian model j. For example, as shown in FIGS. 25A and 25B, a weighting coefficient for cost and a weighting coefficient for risk may be determined in advance according to age. When the age is low or high, the risk weighting coefficient is set to be larger than the cost, and the sum of the weighting coefficient for the cost and the weighting coefficient for the risk may be set to 1. This makes it possible to select the optimal route in consideration of the pedestrian's risk, cost, and individual characteristics, so that the route and behavior of the pedestrian model can be determined based on the route actually selected by the pedestrian. It can be simulated.

  In addition, the case where the optimum route is selected based on the utility calculated by the sum of the cost sum and the risk sum has been described as an example, but the risk sum calculated for each of the route candidates is Based on this, an optimal route may be selected from the route candidates. In this case, a route candidate that minimizes the sum of risks may be selected as an optimal candidate.

  Next, a second embodiment will be described. In the second embodiment, a case where the present invention is applied to a traffic simulation apparatus that simulates the behavior of a pedestrian model that models a pedestrian and a vehicle model that models a vehicle will be described as an example. In addition, about the part which becomes the structure similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted.

  The second embodiment is mainly different from the first embodiment in that the behavior of the vehicle model is simulated together with the behavior of the pedestrian model, and that an accident is caused.

  As shown in FIG. 26, the traffic simulation apparatus 210 according to the second embodiment includes a walking environment database 12, a pedestrian database 14, a driver database 212, an actual accident database 214, a traffic simulation unit 216, And a simulation result database 218.

  The driver database 212 stores driver attributes for driving each vehicle. Specifically, the driver database 212 stores, for each driver, as a driver attribute, a departure place, a destination, a departure time, a route, age, sex, driving skill, and a legal compliance tendency.

  The actual accident database 214 stores driver attributes as perpetrator attributes, pedestrian attributes as victim attributes, accident details, accident factors, accident environments, and accident situations. These data may be input in real time via a network, or may be input by an operator.

  The actual accident database 214 stores the number of accidents for each classified accident content as the accident content. For example, “the number of right-handed accidents due to a recognition error of a straight-ahead vehicle”, “the number of rear-end collisions due to falling asleep”, and “the number of rear-end collisions due to wakimi” are stored. Also, for each accident category, the age, sex, driving experience, victim age, gender, accident-causing factors (conversations with passengers on the front passenger seats, search, mobile phone operation, etc.), accidents The weather at the time of occurrence, the contrast of the accident environment, the size of the intersection, the location of the accident (inside the intersection, immediately before the intersection, etc.), and the accident situation (stopping, following the preceding vehicle, acceleration, etc.) are stored.

  The traffic simulation unit 216 reproduces the behavior of each pedestrian model and each vehicle model based on each data of the walking environment database 12, the pedestrian database 14, the driver database 212, and the actual accident database 214, and traffic Reproduce the flow. The traffic simulation unit 216 includes a road environment route generation unit 218, a road environment update unit 220, a pedestrian behavior calculation unit 222, a vehicle behavior calculation unit 224, and a traffic condition management unit 226.

  The road environment route generation unit 218 creates a walking environment model such as a road model and a building based on the data from the walking environment database 12 and converts the pedestrian model into a road model based on the data from the pedestrian database 14. Place on top. Further, the vehicle model is arranged on the road model based on the data from the driver database 212. Further, based on data from the pedestrian database 14, a plurality of route candidate points and a plurality of links of each pedestrian model are generated on the road model.

  The road environment update unit 220 updates the arrangement position of each vehicle model and each pedestrian model according to the position information and speed information of each vehicle model and each pedestrian model output from the traffic condition management unit 226.

  The pedestrian behavior calculation unit 222 determines an optimum route from route candidates using the route candidate points and links generated by the road environment route generation unit 218. In determining the route, the pedestrian model recognizes the walking environment, surrounding vehicles, and pedestrians according to each pedestrian attribute, determines the optimum route from the route candidate points and the route candidate represented by the link, and selects the desired route. Determine desired behavior such as speed. Then, the behavior is determined based on the determined optimum route and desired behavior. Further, the movement amount is calculated based on the desired speed. Further, based on the data of the actual accident database 214, an error is generated in the behavior of the pedestrian model according to the pedestrian attribute, the road environment to walk, and the walking situation, and an accident is generated.

As in the first embodiment, the route is determined based on the utility U _kj of the pedestrian j for each of the route candidates to the destination (k is a set of links existing on the route candidates). The utility U _kj is calculated for each of the route candidates from the risk and cost according to the above equation (1).

The risk R _ij of the link i of the pedestrian j is a dynamic risk function f _c as a collision risk that changes every moment related to the collision with the vehicle and an environment with danger as shown in the following equation (11). Etc. and a static risk function f _u that does not change.
R _ij = f _c (movement speed, vehicle speed,
Time difference to conflict point, prediction of signal display)
+ _Fu (walking environment, link length, moving speed, age, experience) (11)
here,
f _u = Li / Vij
× (Walking environment weight + α × experience level + β × age level) (12)
f _c = γ × δ × e ^{− (time difference to conflict point)} (13)

  However, γ is a predetermined parameter, and δ is 0 when waiting for the signal to become blue or blue or waiting for the approaching vehicle to pass, and is 1 otherwise.

  As in the above equation (13), the dynamic risk as the collision risk is calculated based on the time difference to the conflict point, and the time difference to the conflict point is the position of the vehicle model relative to the current position of the pedestrian model. And the relative speed between the moving speed of the pedestrian model and the vehicle speed of the vehicle model.

  The vehicle behavior calculation unit 224 determines a desired action (desired route, speed, acceleration / deceleration, etc.) based on recognition according to each driver attribute based on the road environment where the vehicle travels, surrounding vehicles, and surrounding pedestrians. The operation is determined from the desired action, and the behavior of the vehicle is determined based on the determined operation. Further, the movement amount is calculated based on the vehicle and driver attributes. Further, based on the data in the actual accident database 214, an error is generated in the behavior of the vehicle according to the driver attribute, the road environment in which the vehicle travels, and the driving situation, thereby causing an accident.

  The traffic condition management unit 226 manages the position, speed, and acceleration / deceleration of each vehicle model and pedestrian model for each calculation step. It also manages the generation and disappearance of vehicle models and pedestrian models on the road model.

  The simulation result database 218 stores the number of accidents and accident situations obtained as simulation results from the traffic simulation unit 216.

  Next, the operation of the traffic simulation apparatus 210 according to the second embodiment will be described. First, when the operator inputs a simulation start time and an end time to the traffic simulation apparatus 210 and inputs a simulation start instruction, the traffic simulation apparatus 210 executes a simulation processing routine shown in FIG. In addition, about the process similar to 1st Embodiment, the same code | symbol is attached | subjected and detailed description is abbreviate | omitted.

  First, in step 250, a road model is generated based on walking environment information indicating a road map in the walking environment database 12, and a road environment model at a simulation start time is generated based on walking environment information indicating a signal and a building. In step 252, the pedestrian model whose start time corresponds to the departure time is arranged at the departure place on the road model based on the data of the pedestrian database 14, and the start time is determined based on the data of the driver database 212. The vehicle model corresponding to the departure time is arranged at the departure place on the road model.

  In step 104, a plurality of route candidate points and a plurality of links from the departure point to the destination are generated for each pedestrian model.

  In the next step 254, a road environment recognized by the pedestrian model is acquired for the target pedestrian model, and in step 256, a determination process is performed based on the recognized road environment acquired in step 254.

  In the determination process of step 256, the sum of risk and the sum of cost are calculated for each candidate route based on the recognized road environment, and the utility is calculated based on the sum of risk and the sum of cost. . Then, a route candidate having the minimum utility is searched, and the searched route candidate is determined as the optimum route.

  In step 257, the behavior of the target pedestrian model is calculated based on the determined optimum route. In step 257, based on the data in the actual accident database 214, an error is generated in the behavior of the pedestrian model according to the pedestrian attributes and the walking road environment recognized in step 254. generate.

  In the next step 112, it is determined whether or not the processing of steps 254 to 257 has been executed for all the pedestrian models that are arranged. Return to, and perform recognition, judgment, and behavior processing for the next pedestrian model. On the other hand, when the processes of steps 254 to 257 are executed for all pedestrian models, the process proceeds to step 258.

  In step 258, the driving road environment recognized by the driver is acquired for the driver of the target vehicle model. In step 260, the desired route, the desired speed, and the desired road environment are recognized based on the driving road environment recognized in step 258. Determine desired behavior such as acceleration / deceleration. In step 262, the operation of the vehicle model is determined based on the desired behavior determined in step 260. In step 264, the behavior of the target vehicle model is determined based on the operation determined in step 262. Calculate

  In step 264, an error occurs in the behavior of the vehicle model in accordance with the driver attribute and the road environment recognized in step 258 based on the data in the actual accident database 214, thereby generating an accident. .

  In the next step 266, it is determined whether or not the processes in steps 258 to 264 have been executed for all the vehicle model drivers arranged. If not, the process goes to step 258. Returning, recognition, judgment, operation, and behavior processing are executed for the driver of the next vehicle model. On the other hand, when the processes of steps 259 to 264 are executed for all the vehicle model drivers, the process proceeds to step 114.

  In step 114, the current time is updated to the next time. In step 268, the road environment model at the updated time is updated, and the behavior of the pedestrian model calculated in step 110 and the calculation in step 264 are performed. The position of the pedestrian model and the vehicle model is updated based on the behavior of the vehicle model. Note that the pedestrian model and the vehicle model that have arrived at the destination are extinguished at this point.

  In the next step 118, it is determined whether or not it is an end time. If the current time is the end time, the simulation processing routine is ended. If the current time is not the end time, in step 270, It is determined whether or not a vehicle model is to be added, and if there is no driver whose current time is the departure time based on the data in the driver database 212, the process proceeds to step 120, while the current time is the departure time. If a driver is present, it is determined that a vehicle model is to be added, and in step 272, the driver's vehicle model whose current time is the departure time is placed at the departure point on the road model, and the process proceeds to step 120. .

  In step 120, it is determined whether or not a pedestrian model is to be added, and if there is no pedestrian model whose current time is the departure time, the process returns to step 254, while walking whose current time is the departure time. If a pedestrian model exists, it is determined that a pedestrian model is to be added, and in step 122, a pedestrian model whose current time is the departure time is placed at the departure place on the road model.

  In step 124, a plurality of route candidate points and a plurality of links from the departure point to the destination are generated for each of the pedestrian models newly arranged in step 122, and the process returns to step 254.

  By executing the simulation processing routine described above, based on the actual accident data, according to the pedestrian attribute and the walking environment, an error occurs in the behavior of the pedestrian model, and according to the driver attribute and the driving environment, An error occurs in the behavior of the vehicle model, causing a traffic accident. Thus, the occurrence of an accident according to the environment of the pedestrian model and the vehicle model can be simulated, and information indicating the number of accidents and the accident situation can be acquired as a simulation result.

  As described above, according to the driving support apparatus according to the second embodiment, in addition to the static risk, the dynamic risk indicating the collision risk with the vehicle traveling near the current position of the pedestrian is considered. Thus, the optimum route can be selected from the route candidates.

  In addition to the behavior of the pedestrian model, the behavior of the vehicle model and the occurrence of an accident can be simulated.

  In addition, although the case where the optimal route is selected based on the utility calculated by the sum of the cost sum and the risk sum has been described as an example, based on the risk sum, the optimal route candidate is selected. A route may be selected. In this case, a route candidate that minimizes the sum of the sum of static risks and the sum of dynamic risks may be selected as the optimum candidate.

  Next, a traffic simulation apparatus according to the third embodiment will be described. In addition, about the part which becomes the same structure as 1st Embodiment and 2nd Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted.

  The third embodiment is different from the second embodiment in that a traffic simulation is performed for each of the presence or absence of the safety system and the effect of the safety system is evaluated.

  As shown in FIG. 28, the traffic simulation apparatus 310 according to the third embodiment includes a walking environment database 12, a pedestrian database 14, a driver database 212, an actual accident database 214, a traffic simulation unit 316, A simulation result database 318 and an effect evaluation unit 330 are provided.

  The traffic simulation unit 316 includes a road environment route generation unit 218, a road environment update unit 220, a safety system unit 320, a pedestrian behavior calculation unit 322, a vehicle behavior calculation unit 324, and a traffic condition management unit 226. ing.

  The safety system unit 320 obtains the situation of the road environment model, the vehicle model, and the pedestrian model from the road environment update unit 220, determines the risk level of the vehicle model or the pedestrian model, and then creates a dangerous vehicle model or pedestrian model. Provide and control danger information. If the vehicle model is equipped with a safety system based on settings from the operator, the vehicle model is provided and controlled with danger information, and if the pedestrian model carries the safety system, walking Provide and control danger information for the person model. Further, if the safety system is not mounted on the vehicle model based on the setting from the operator and the pedestrian model does not carry the safety system, no danger information is provided or controlled.

  In the process of recognition and determination, the pedestrian behavior calculation unit 322 determines an optimum route from a plurality of route candidate points generated by the road environment route generation unit 218 and route candidates represented using a plurality of links. Then, the behavior is calculated based on the determined route. Further, based on the data of the actual accident database 214, an error is generated in the behavior of the pedestrian model according to the pedestrian attribute, the road environment to walk, and the walking situation, and an accident is generated. For a pedestrian model carrying a safety system, risk information and control are obtained from the safety system unit 320, and judgment and behavior based on the risk information and control are determined.

  The vehicle behavior calculation unit 324 determines the desired action (desired route, speed, acceleration / deceleration, etc.) based on the recognition according to each driver attribute based on the road environment, the surrounding vehicle, and the surrounding pedestrians. And determining the behavior of the vehicle model based on the operation. Further, based on the data in the actual accident database 214, an error is generated in the behavior of the vehicle model in accordance with the driver attribute, the driving environment, and the driving situation, thereby generating an accident. For the driver of the vehicle model equipped with the in-vehicle device of the safety system, the danger information and the vehicle control are obtained from the safety system unit 320, and the judgment, operation, or behavior based on the danger information and control is determined.

  The simulation result database 318 stores the number of accidents and accident situations obtained as simulation results from the traffic simulation unit 316 and stores the accident reduction rate by the safety system output from the effect evaluation unit 330.

  The effect evaluation unit 330 obtains, from the simulation result database 318, the number of accidents when the safety system is not used and the number of accidents when the safety system is used, and calculates and outputs the accident reduction rate by the safety system. To do.

  Next, the operation of the traffic simulation apparatus 310 according to the third embodiment will be described.

  According to the input of the operator, the safety system is set not to be mounted on the vehicle model and to be carried by the pedestrian model, and the simulation processing routine is executed as in the second embodiment, and the simulation result The number of accidents and the situation of the accidents obtained as the output of are stored in the simulation result database 318.

  Next, the vehicle model is set to be equipped with a safety system, and a simulation processing routine is executed. In this simulation processing routine, in accordance with the provision and control of the danger information obtained from the safety system unit 320, each process of recognition, judgment, operation, and behavior for the driver of the vehicle model equipped with the in-vehicle device of the safety system is performed. Execute. Then, the number of accidents and the accident situation obtained as simulation result output are stored in the simulation result database 318.

  And the accident reduction rate by a safety system is calculated, and it outputs as evaluation of the effect of the safety system of onboard equipment.

  Next, the pedestrian model is set to carry the safety system, and the simulation processing routine is executed. In this simulation processing routine, each process of recognition, determination, and behavior for the pedestrian model carrying the safety system is executed in accordance with the provision and control of the danger information obtained from the safety system unit 320. Then, the number of accidents and the accident situation obtained as simulation result output are stored in the simulation result database 318.

  And the accident reduction rate by a safety system is calculated, and it outputs as an evaluation of the effect of the safety system for pedestrian carrying.

  As described above, according to the driving support apparatus according to the third embodiment, the effect of the safety system mounted on the vehicle and the effect of the safety system carried by the pedestrian can be evaluated.

  Next, a fourth embodiment will be described. In the fourth embodiment, a case where the present invention is applied to a pedestrian behavior prediction apparatus that predicts a pedestrian behavior will be described as an example.

  In the pedestrian behavior predicting apparatus according to the fourth embodiment, the current position is input as the departure point for the pedestrian to be predicted, and the destination is estimated based on the pedestrian situation.

  Then, based on the road map information prepared in advance and the start point and destination of the pedestrian to be predicted, a plurality of route candidate points and a plurality of links are generated on the road map, A plurality of route candidates represented by

  Then, for each of a plurality of route candidates, the type of moving location corresponding to the link existing on the route candidate is considered using the same method as in the first embodiment or the second embodiment. Then, calculate the sum of the risk and the sum of the cost, and calculate the utility. Search for a candidate for the route with the smallest calculated utility, select the candidate for the searched route as the optimum route, calculate the behavior of the pedestrian to be predicted based on the selected route, and predict Output as the action to be performed.

  As described above, according to the pedestrian behavior prediction device according to the fourth embodiment, for each of the route candidates, the type of the moving location corresponding to the link existing on the route candidate is considered. Predict the pedestrian's behavior with high accuracy by calculating the sum and the sum of the costs, considering the risk and cost of each candidate route, selecting the route and predicting the pedestrian's behavior Can do.

  Next, a fifth embodiment will be described. In the fifth embodiment, a case where the present invention is applied to a route search device that searches for an optimum route and displays it to an operator will be described as an example.

  In the route search device according to the fifth embodiment, a departure place and a destination are input from an operator who is a pedestrian. In addition, about the departure place, you may make it input the present position acquired using GPS etc. as a departure place.

  Then, based on the road map information prepared in advance and the pedestrian's departure place and destination, a plurality of route candidate points and a plurality of links are generated on the road map, and are represented by the route candidate points and the links. A plurality of route candidates to be generated are generated.

  Then, for each of a plurality of route candidates, the type of moving location corresponding to the link existing on the route candidate is considered using the same method as in the first embodiment and the second embodiment. Then, calculate the sum of the risk and the sum of the cost, and calculate the utility. The candidate of the route with the smallest calculated utility is searched, the searched route candidate is selected as the optimum route, and the selected route is displayed.

  As described above, according to the route search device according to the fifth embodiment, for each of the route candidates, the sum of the risk and the cost according to the type of the moving location corresponding to the link existing on the route candidate. The optimal route can be searched for by selecting the route in consideration of the risk and cost of each of the route candidates.

  In the above embodiment, the case where the selected optimum route is displayed has been described as an example. However, the present invention is not limited to this, and each of the generated route candidates and each route candidate are displayed. It is also possible to display the utility (or the sum of risk and the sum of cost) calculated for the operator and allow the operator to select a desired route.

  Next, a sixth embodiment will be described. The sixth embodiment is different from the fifth embodiment in that the utility considering the risk and cost is calculated and displayed for the route to each waypoint.

  In the route search device according to the sixth embodiment, a current position and a destination are input from an operator who is a pedestrian. The current position may be acquired using GPS or the like.

  Then, based on the road map information prepared in advance and the current position and destination of the pedestrian, a plurality of route candidate points and a plurality of links are generated on the road map, and are generated as route candidates. A route to each of a plurality of route candidate points is generated.

  Then, for each of the plurality of generated routes, using the same method as in the first and second embodiments, the type of moving location corresponding to the link existing on the route is considered. Calculate the utility by calculating the sum of the risk and the sum of the costs. Then, each of the generated route candidate points and the utility (or the sum of risk and the sum of costs) calculated for the route to each route candidate point are displayed. Then, the operator selects a desired route based on the displayed utility.

  As described above, according to the route search device according to the sixth embodiment, for each of the routes up to the route candidate points, the sum of the risk and the sum of the risks are determined according to the type of moving location corresponding to the link existing on the route. The operator can select a desired route by calculating the sum of the costs and displaying the utility considering the risk and cost for each route to the route candidate point.

In the above embodiment, the case where the route search is performed using the Dikstra method has been described as an example. However, the present invention is not limited to this, and the A ^* method, CA method, and neural network are used for route determination. Other route searching methods such as a method may be used.

  In the above description, the prediction target is a pedestrian. However, the prediction target is not limited to the pedestrian but may be a bicycle or a wheelchair. Furthermore, a motorcycle may be applied as the vehicle.

It is a block diagram which shows the pedestrian simulation apparatus which concerns on the 1st Embodiment of this invention. (A) It is an image figure which shows a mode that the route candidate point was produced | generated on the road model, (B) It is an image figure which shows a mode that the link was produced | generated on the road model. (A) It is an image figure which shows a mode that a route candidate point was produced | generated at the signalized intersection on a road model, (B) It is an image figure which shows a mode that the link was produced | generated at the signalized intersection on a road model. It is an image figure which shows a mode that a route candidate point and a link were produced | generated when there exists an obstruction on a road model. It is an image figure which shows a mode that a route candidate point and a link were produced | generated at the signalized intersection on a road model. It is a figure for demonstrating that two links exist in the produced | generated link. It is a graph which shows the relationship between age and each of walking speed and running speed. It is a graph which shows the relationship between the consciousness level and the consciousness level level regarding a moving speed. It is a graph which shows the relationship between a physical ability and the physical ability level regarding a moving speed. It is a graph which shows the relationship between a consciousness level and the consciousness level level regarding a prediction error. It is a graph which shows the relationship between age and the age level regarding a prediction error. It is a graph which shows the relationship between experience and the experience level regarding a prediction error. It is a table | surface which shows the example of the weighting coefficient of the walking environmental load with respect to the kind of moving place. It is a graph which shows the relationship between a physical ability and the physical ability level regarding a walking environmental load. It is a graph which shows the relationship between age and the age level regarding walking environmental load. It is a table | surface which shows the example of the weighting coefficient of the walking environmental load with respect to the kind of weather. It is a graph which shows the relationship between the physical ability level regarding the physical ability and the running load. It is a graph which shows the relationship between age and the age level regarding the load to run. It is a table | surface which shows the example of the weighting coefficient of the violation action load with respect to the kind of violation content. It is a graph which shows the relationship between a law compliance tendency and a law compliance level. It is a table | surface which shows the example of the weighting coefficient of the risk with respect to the kind of movement place. It is a graph which shows the relationship between experience and the experience level regarding a risk. It is a graph which shows the relationship between age and the age level regarding a risk. It is a flowchart which shows the content of the simulation process routine in the pedestrian simulation apparatus which concerns on the 1st Embodiment of this invention. (A) A table showing an example of a weighting factor for a cost with respect to age and a weighting factor for a risk, and (B) a graph showing a relationship between each of a weighting factor with respect to age and a cost and a weighting factor with respect to a risk. It is a block diagram which shows the traffic simulation apparatus which concerns on the 2nd Embodiment of this invention. It is a flowchart which shows the content of the simulation process routine in the traffic simulation apparatus which concerns on the 2nd Embodiment of this invention. It is a block diagram which shows the traffic simulation apparatus which concerns on the 3rd Embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Pedestrian simulation apparatus 12 Walking environment database 14 Pedestrian database 16 Pedestrian simulation part 18 Walking environment creation update part 20 Via candidate point link production | generation part 22,222,322 Pedestrian behavior calculation part 210,310 Traffic simulation apparatus 212 Driver database 214 Actual accident database 216, 316 Traffic simulation unit 218, 318 Simulation result database 218 Road environment route generation unit 220 Road environment update unit 224, 324 Vehicle behavior calculation unit 226 Traffic condition management unit 320 Safety system unit 330 Effect evaluation unit

Claims (18)

A route candidate generating means for generating a plurality of route candidates from the starting point of the pedestrian or bicycle to the destination based on the map information;
On each of the route candidates generated by the route generation means, based on the risk level when moving in advance for each of a plurality of types of pedestrians or bicycles, on the route candidates. Calculating means for calculating the total risk according to the type of the moving location existing in
Output means for outputting the total risk of each of the route candidates calculated by the calculation means;
A route search device including:   The route search apparatus according to claim 1, wherein the output unit displays a plurality of route candidates generated by the route generation unit and a total risk of each of the route candidates.   Route selection means for selecting a route of the pedestrian or bicycle from a plurality of route candidates generated by the route generation means based on the total risk of each of the route candidates output by the output means. The route search device according to claim 1, further comprising:   The route search according to any one of claims 1 to 3, wherein the plurality of types of pedestrians or bicycles include a pedestrian crossing, a sidewalk separated by pedestrians, a sidewalk not separated by pedestrians, and a roadway. apparatus.   The route search device according to claim 4, wherein a risk degree for the roadway is determined for each of vehicle traffic per unit time of the roadway or each of a plurality of average vehicle speeds of a vehicle traveling on the roadway. The pedestrian collides with the vehicle for each of the route candidates generated by the route generation means based on at least one of the position of the vehicle with respect to the current position of the pedestrian or bicycle and the speed of the vehicle. A collision risk calculation means for calculating the collision risk,
A plurality of route candidates generated by the route generation unit based on the total risk level of each of the route candidates output by the output unit and the collision risk level calculated by the collision risk level calculation unit. The route search device according to claim 1, further comprising route selection means for selecting a route of the pedestrian or the bicycle. The route candidate generation means generates a plurality of route candidate points that pass through the route from the departure point to the destination and a plurality of links that connect two route candidate points of the plurality of route candidate points. , Generate a plurality of candidate routes using the generated links,
The calculation means calculates the risk for each of the generated plurality of links according to the type of the moving location existing on the link, and based on the calculated risk of each of the plurality of links. Then, for each of the route candidates generated by the route generation means, the total risk according to the type of the moving location existing on the route candidate is calculated. The route search device according to item. The travel time is calculated for each of the route candidates generated by the route candidate generating means, and the load amount when moving in advance is determined according to each of a plurality of types of moving places of the pedestrian or bicycle. Based on the calculated travel time and total load amount, based on the calculated travel time and the total load amount for each of the plurality of route candidates, For each of a plurality of route candidates, cost calculating means for calculating the cost when moving,
Route selection for selecting the route of the pedestrian or the bicycle from a plurality of route candidates generated by the route generation unit based on the risk level output by the output unit and the cost calculated by the cost calculation unit Means,
The route search device according to any one of claims 1 to 7, further comprising:   The route search device according to claim 8, wherein the total load amount is calculated based on the load amount, weather, whether to run and move, and whether the movement is in violation of the law.   The route selection means is based on the total risk degree output by the output means, the cost calculated by the cost calculation means, and the personal characteristics of the set pedestrian or bicycle, based on the characteristics of the pedestrian or bicycle. The route search device according to claim 8 or 9, wherein a route is selected. Based on the map information, a waypoint generating means for generating a plurality of waypoints present on the route candidate from the current position of the pedestrian or bicycle to the destination,
Based on the risk when moving in advance for each of a plurality of types of pedestrians or bicycle movement locations, for each of the waypoints generated by the waypoint generation means, from the current position A calculating means for calculating a total risk according to the type of the moving location existing on the route to the waypoint;
Output means for outputting the total risk of each of the waypoints calculated by the calculation means;
A route search device including: A route search device according to claim 3, 6, 8, 9, or 10;
A vehicle model that models a vehicle is run, and a pedestrian model that models a pedestrian or a bicycle model that models a bicycle is moved in accordance with the route selected by the route search device to simulate a traffic state. Simulation means;
Including traffic simulation equipment. For each driver attribute and accident environment, driver accident information indicating a driver or vehicle event that causes an accident is stored, and for each pedestrian attribute or bicycle attribute and each accident environment, a pedestrian or bicycle that causes an accident. Storage means for storing pedestrian accident information representing the events of
A driver event generating means for generating a driver or vehicle event as an accident inducing factor based on the driver accident information stored in the storage means;
Pedestrian event generating means for generating a pedestrian or bicycle event that becomes an accident inducing factor based on the pedestrian accident information stored in the storage means,
The simulation means generates the event when a driver or vehicle event that causes an accident is generated based on the driver accident information stored in the storage means and the road environment in which the vehicle model travels. Vehicle behavior determining means for determining the behavior of the vehicle model, pedestrian accident information stored in the storage means, the route selected by the route search device, and the road environment or bicycle on which the pedestrian model walks When a pedestrian or bicycle event that causes an accident is generated based on the road environment in which the model travels, the pedestrian who determines the behavior of the pedestrian model or the bicycle model so as to generate the event Behavior determining means, and when the vehicle model is driven based on the behavior of the vehicle model determined by the vehicle behavior determining means, , Based on the behavior of the pedestrian model or bicycle model was determined by the pedestrian behavior determining means, said moving the pedestrian model or bicycle model, traffic simulation apparatus according to claim 12, wherein simulating the traffic conditions. Evaluation means for evaluating the effect of the safety system by comparing the result of simulating the traffic state with the safety system mounted on the vehicle model and the result of simulating the traffic state without mounting the safety system on the vehicle model Further including
The vehicle behavior determining means determines the behavior of the vehicle model based on driver accident information stored in the storage means, a road environment in which the vehicle model travels, and whether a vehicle model safety system is installed. Traffic simulation equipment. By comparing the result of simulating the traffic state with the pedestrian model or bicycle model carrying the safety system and the result of simulating the traffic state without carrying the safety system to the pedestrian model or bicycle model, the safety An evaluation means for evaluating the effectiveness of the system;
The pedestrian behavior determination means includes pedestrian accident information stored in the storage means, a route selected by the route search device, a road environment where a pedestrian model walks or a road environment where a bicycle model runs, and a pedestrian The traffic simulation device according to claim 13 or 14, wherein the behavior of the pedestrian model or the bicycle model is determined based on whether the model or the bicycle model safety system is carried. A route search device according to claim 3, 6, 8, 9, or 10;
A pedestrian behavior prediction device including prediction means for predicting the behavior of a pedestrian or bicycle to be predicted based on the route selected by the route search device. Computer
A route candidate generating means for generating a plurality of route candidates from the starting point of the pedestrian or bicycle to the destination based on the map information;
On each of the route candidates generated by the route generation means on the basis of the risk level when moving in advance for each of a plurality of types of pedestrians or bicycles, on the route candidates. A program for functioning as a calculation means for calculating the total risk according to the type of the moving location existing in the vehicle, and an output means for outputting the total risk of each of the route candidates calculated by the calculation means. Computer
A route candidate generating means for generating a plurality of route candidates from the starting point of the pedestrian or bicycle to the destination based on the map information;
On each of the route candidates generated by the route generation means on the basis of the risk level when moving in advance for each of a plurality of types of pedestrians or bicycles, on the route candidates. Calculating means for calculating the total risk according to the type of the moving location existing in
Output means for outputting the total risk of each of the route candidates calculated by the calculation means;
A route selection unit that selects a route of the pedestrian or the bicycle from a plurality of route candidates generated by the route generation unit based on the total risk of each of the route candidates output by the output unit; And driving a vehicle model that models a vehicle and moving a pedestrian model that models a pedestrian or a bicycle model that models a bicycle according to the route selected by the route selection means to simulate traffic conditions A program for functioning as a simulation means.

Images