JP2007230454A - Course setting method, device, and automatic driving system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a course setting method, device and program, and an automatic driving system, in which safety can be ensured even in an actually possible situation. <P>SOLUTION: A computer including a storage means storing at least positions of a plurality of objects and an internal state of each object including its speed reads the positions of the plurality of objects and the internal states from the storage means, generates, based on the read object positions and internal states, a change in possible position each of the plurality of objects can take with the lapse of time as a trace on a time space composed of time and space, performs stochastic prediction of course for each of the plurality of objects by use of the generated trace, calculates, based on the predicted result, an interference degree quantitatively showing the degree of interference of the possible course of a specific object among the plurality of objects with each possible course of the other objects, and selects a course to be taken by the specific object according to the calculated interference degrees. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a course setting method, apparatus, and program for setting a course to be taken by a specific object by predicting a course that can be taken by a specific object included in a plurality of objects and a course that can be taken by another object. And an automatic driving system.

  In recent years, various attempts have been made to realize automatic driving of a moving body such as a four-wheeled vehicle. In order to realize automatic driving of moving objects, it is important to accurately detect objects such as vehicles, pedestrians, and obstacles around the vehicle, and to avoid danger during traveling based on the detection results. . Among these, object detection techniques using various sensors and various radars are known as techniques for accurately detecting surrounding objects.

  The automatic driving technique for a moving body is a technique for automatically moving a moving body from a departure place to a destination when a destination is input. When the movement range is narrow, this technique can be reduced to a route search technique by creating a map of the movement range in advance and predicting the influence of a dynamic obstacle in advance. However, in the case where the moving range of the moving body is wide, such as when the moving body is an automobile, the automatic driving technique cannot be reduced to the route search technique. The wide range here refers to a range in which the time t necessary for avoiding a dynamic obstacle and the time τ necessary for traveling through the entire process are remarkably different. For example, t is about several seconds. On the other hand, τ is about several hours.

  There are two main reasons why the automatic driving technique cannot be reduced to the route search technique when the moving range of the moving body is wide. First, the first reason will be described. Consider a situation after about 10 tons of time has elapsed since the mobile body departed from the departure place. In this case, the influence of a dynamic obstacle spreads over the entire road, and a route that does not collide cannot be defined. That is, when the moving range of the moving body is wide, it is impossible to calculate the route from the starting point to the destination in advance. Next, the second reason will be described. When the moving range of the moving body is wide, as described above, the time τ required to travel through the entire process is very long compared to t. For this reason, it is impossible for a computer mounted on an automobile to finish a necessary calculation within a practical time in which actual collision avoidance can be realized.

  As described above, in the automatic driving technology of a moving body that moves over a wide area such as an automobile, at least the influence of other dynamic obstacles is not taken into consideration, or in addition to the route search technique that does not require the influence in practice. Thus, a route calculation technique for calculating a route for ending the calculation necessary for avoiding a collision with a dynamic obstacle in a practical time and avoiding danger during traveling is required.

  As the above-mentioned route calculation technique, as a technique for avoiding danger during traveling, in a system composed of a plurality of objects and the own vehicle, information on the position and speed of the own vehicle and positions of a plurality of objects other than the own vehicle And the information on the speed are used to generate the course of each object including the vehicle and predict the possibility that any two of the objects constituting the system will collide (for example, , See Non-Patent Document 1). In this technology, the paths that can be taken by all the objects constituting the system are predicted and output by an operation sequence of the same framework using the probability concept. Then, based on the obtained prediction result, a route that realizes the safest situation for the entire system including the vehicle is obtained and output.

A. Broadhurst, S. Baker, and T. Kanade, "Monte Carlo Road Safety Reasoning", IEEE Intelligent Vehicle Symposium (IV2005), IEEE, (June 2005)

  However, the technique disclosed in Non-Patent Document 1 mainly focuses on predicting a course in which all the objects constituting the system are safe, and the course obtained by such prediction is, It is not clear whether the safety for a specific object (for example, the vehicle) is sufficiently secured.

  This point will be described more specifically. In an actual road situation, a driver or a pedestrian of another vehicle may make a mistake in recognizing the road situation, and may exhibit undesirable behavior for surrounding objects including the own vehicle without his / her awareness. On the other hand, in the non-patent document 1 described above, since it is implicitly assumed that all objects exhibit behavior giving priority to safety, a certain object exhibits undesirable behavior for surrounding objects. Thus, it was unclear whether sufficient safety could be ensured even in situations that could occur in reality.

  The present invention has been made in view of the above, and it is an object of the present invention to provide a course setting method, apparatus, program, and automatic driving system that can ensure safety even in a situation that can occur in reality. And

  In order to solve the above-mentioned problems and achieve the object, the invention according to claim 1 is a computer comprising storage means for storing at least the positions of a plurality of objects and an internal state including the speed of each object. A path setting method for setting a path to be taken by the specific object by predicting a path that can be taken by a specific object included in the plurality of objects and a path that can be taken by the other object. When the position and the internal state are read from the storage means, and based on the read position and internal state of the object, a change in position that each of the plurality of objects can take with time is configured from time and space. A trajectory generation step generated as a trajectory in space, and the path probabilities of the plurality of objects by using the trajectory generated in the trajectory generation step An interference level that quantitatively indicates a degree of interference between a course that can be taken by the specific object and a course that can be taken by the other object, based on a prediction step that performs each prediction, and a result predicted by the prediction step. An interference degree calculating step for calculating, and a route selecting step for selecting a route to be taken by the specific object in accordance with the interference degree calculated in the interference degree calculating step.

  According to a second aspect of the present invention, in the first aspect of the invention, the degree of interference has a smaller value as the degree of interference between the path that the specific object can take and the path that the other object can take. In the course selection step, the course having the minimum degree of interference is selected.

  The invention according to claim 3 is the invention according to claim 2, wherein when there are a plurality of paths with the minimum degree of interference, the path selection step is most suitable for a predetermined additional selection criterion from among the plurality of paths. It is characterized by selecting a course.

  According to a fourth aspect of the present invention, in the first aspect of the invention, the degree of interference has a larger value as the degree of interference between the path that can be taken by the specific object and the path that can be taken by the other object is smaller. In the course selection step, the course having the maximum degree of interference is selected.

  According to a fifth aspect of the present invention, in the fourth aspect of the present invention, when there are a plurality of paths having the maximum degree of interference, the path selection step is most suitable for a predetermined additional selection criterion among the plurality of paths. It is characterized by selecting a course.

  According to a sixth aspect of the present invention, in the invention according to any one of the second to fifth aspects, the operation signal corresponding to the history of the position of the route selected in the route selection step and the operation sequence realizing the route is generated. The operation signal transmitting step of transmitting the generated operation signal to the outside is further provided.

  According to a seventh aspect of the present invention, in the invention according to any one of the first to sixth aspects, the interference degree calculating step is a spatial mode in which the specific object and each of the other objects interfere with each other. The degree of interference between the specific object and each of the other objects is increased or decreased by a predetermined amount according to the number of times of approaching the interference distance, which is a distance.

  The invention according to claim 8 is the invention according to any one of claims 1 to 7, wherein the interference degree calculating step weights each interference degree value between the specific object and the other object. It is characterized by taking the sum.

  The invention according to claim 9 is the invention according to any one of claims 1 to 8, wherein the trajectory generation step includes an operation selection step for selecting an operation for the object from a plurality of operations, and the operation selection step. An object operation step for operating the selected operation for a predetermined time, and a position and an internal state of the object after operating the selected operation in the object operation step are control conditions relating to the control of the object and a movable area of the object A determination step of determining whether or not a movement condition related to is satisfied, and a series of processing from the operation selection step to the determination step is repeatedly performed until a trajectory generation time for generating a trajectory is reached. To do.

  The invention described in claim 10 is characterized in that in the invention described in any one of claims 1-9, there is further provided an output step of outputting information on the route selected in the route selection step.

  The invention according to claim 11 is a course setting device that sets a course to be taken by the specific object by predicting a course that can be taken by the specific object included in the plurality of objects and a course that can be taken by the other object. A storage unit for storing at least the positions of the plurality of objects and an internal state including a speed of each object; and the positions and the internal states of the plurality of objects are read from the storage unit; And a trajectory generating means for generating a change in position that each of the plurality of objects can take with the passage of time as a trajectory on time and space composed of time and space, based on the internal state, and the trajectory generating means And a prediction unit that performs probabilistic prediction of the courses of the plurality of objects by using the trajectory generated in the step, and based on a result predicted by the prediction unit Interference degree calculating means for calculating an interference degree quantitatively indicating the degree of interference between the course that can be taken by the specific object and the course that can be taken by the other object, and depending on the interference degree calculated by the interference degree calculating means Route selection means for selecting a route to be taken by the specific object.

  According to a twelfth aspect of the present invention, in the invention of the eleventh aspect, the degree of interference has a smaller value as the degree of interference between the path that the specific object can take and the path that the other object can take. The route selection means selects the route with the minimum degree of interference.

  According to a thirteenth aspect of the present invention, in the invention according to the twelfth aspect, when there are a plurality of paths with the minimum degree of interference, the path selection means is most suitable for a predetermined additional selection criterion among the plurality of paths. It is characterized by selecting a course.

  According to a fourteenth aspect of the invention, in the invention of the eleventh aspect, the degree of interference has a larger value as the degree of interference between the path that the specific object can take and the path that the other object can take. The route selecting means selects the route having the maximum degree of interference.

  The invention according to claim 15 is the invention according to claim 14, wherein when there are a plurality of routes having the maximum degree of interference, the route selecting means is most suitable for a predetermined additional selection criterion among the plurality of routes. It is characterized by selecting a course.

  A sixteenth aspect of the present invention is the method according to any one of the twelfth to fifteenth aspects, wherein the operation signal corresponding to the history of the position of the route selected by the route selecting means and the operation sequence that realizes the route is generated. The operation signal transmitting means for transmitting the generated operation signal to the outside is further provided.

  According to a seventeenth aspect of the present invention, in the invention according to any one of the eleventh to sixteenth aspects, the interference degree calculating unit is a spatial unit in which the specific object and each of the other objects interfere with each other. The degree of interference between the specific object and each of the other objects is increased or decreased by a predetermined amount according to the number of times of approaching the interference distance, which is a distance.

  The invention according to claim 18 is the invention according to any one of claims 11 to 17, wherein the interference degree calculating means weights each interference degree value between the specific object and the other object. It is characterized by taking the sum.

  The invention according to claim 19 is the invention according to any one of claims 11 to 18, wherein the trajectory generating means includes an operation selecting means for selecting an operation for the object from a plurality of operations, and the operation selecting means. Object operating means for operating the selected operation for a predetermined time, and the position and internal state of the object after the selected operation is operated by the object operating means are control conditions relating to the control of the object and the movable area of the object And a determination unit that determines whether or not the movement condition is satisfied. A series of processes from the operation selection process by the operation selection unit to the determination process by the determination unit reaches a locus generation time for generating a locus. It is characterized by being repeatedly performed.

  A twentieth aspect of the invention is characterized in that in the invention of any one of the eleventh to nineteenth aspects, the apparatus further comprises output means for outputting information relating to the course selected by the course selection means.

  According to a twenty-first aspect of the present invention, a course setting program causes the computer to execute the course setting method according to any one of the first to tenth aspects.

  The invention according to claim 22 is an automatic driving system that is mounted on a vehicle and automatically drives the vehicle, wherein the route setting device according to any one of claims 11 to 20 and the route setting device include: And an actuator device for operating the vehicle in response to an operation signal for realizing the route selected by the route selecting means.

  According to the present invention, a computer having storage means for storing at least the positions of a plurality of objects and the internal state including the speed of each object reads out the positions and internal states of the plurality of objects from the storage means. Based on the read position and internal state of the object, a change in position that each of the plurality of objects can take as time elapses is generated as a trajectory on time and space composed of time and space. The paths of the plurality of objects are probabilistically predicted by using the trajectories, and based on the predicted results, interference between the paths that can be taken by the specific object and the paths that can be taken by the other objects is determined. It occurs as a reality by calculating the degree of interference that quantitatively indicates the degree, and selecting the course that the specific object should take according to the calculated degree of interference. Also it is possible to achieve a secure safety under that situation.

  The best mode for carrying out the present invention (hereinafter referred to as “embodiment”) will be described below with reference to the accompanying drawings.

(Embodiment 1)
FIG. 1 is a block diagram showing a functional configuration of the automatic driving system according to Embodiment 1 of the present invention. The automatic driving system 100 shown in the figure is mounted on a moving body such as a four-wheeled vehicle, and can take a course that can be taken by the own vehicle as a specific object and other objects (including vehicles, people, obstacles, etc.). A route setting device 1 that sets a route to be taken by the vehicle by predicting the route and an actuator device 11 that operates the vehicle according to an operation signal that realizes the route set by the route setting device 1 are provided. .

  The course setting device 1 includes an input unit 2 to which various information is input from the outside, a sensor unit 3 that detects the position and internal state of an object existing in a predetermined range, and a result detected by the sensor unit 3. A trajectory generation unit 4 that generates a change in the position that an object can take as time passes as a trajectory in space-time composed of time and space, and the probability of the course of the object using the trajectory generated by the trajectory generation unit 4 Interference for quantitatively indicating the degree of interference between the course that can be taken by the vehicle and the course that can be taken by other objects based on the result predicted by the prediction section 5 A degree calculation unit 6, a route selection unit 7 that selects a route to be taken by the vehicle according to the interference degree calculated by the interference degree calculation unit 6, and an operation signal corresponding to the selection result of the route selection unit 7. Operation signal transmission to the actuator device 11 It comprises 8, an output unit 9 for outputting various information about the processing performed by the route setting device 1, a storage unit 10 for storing various information including the position and the internal state of an object detected by the sensor unit 3, a.

  The input unit 2 has a function of inputting various setting information and the like when predicting the course of an object, and includes a remote controller, a keyboard (including a touch panel format that can be input on the screen), a pointing device (mouse, For example, a trackpad). Further, a microphone capable of inputting information by voice may be provided as the input unit 2. When various setting information is determined in advance, the function of the input unit 2 may be replaced by a storage unit 10 having a ROM (Read Only Memory) or the like that stores them.

  The sensor unit 3 is realized by using a millimeter wave radar, a laser radar, an image sensor, or the like. The sensor unit 3 includes various sensors such as a speed sensor, an acceleration sensor, a rudder angle sensor, and an angular velocity sensor, and can also detect the movement state of the host vehicle. Note that the internal state of the object detected by the sensor unit 3 is a useful state that can be used for prediction of the object, and preferably the speed (with speed and direction) and angular speed (size and size) of the object. Physical quantity). In addition, the case where the value of those physical quantities is 0 (a state where the object is stopped) is also included.

  The trajectory generation unit 4 predicts and generates a trajectory that an object can take until a predetermined time elapses, and an operation selection unit that selects an operation for virtually operating the object on a simulation from a plurality of operations. 41, an object operation unit 42 for performing the operation selected by the operation selection unit 41 for a predetermined time, and determining whether or not the position and internal state of the object after the operation by the object operation unit 42 satisfy a predetermined condition And a determination unit 43.

  The output unit 9 includes a display unit 91 that displays and outputs information related to the processes performed by the prediction unit 5, the interference degree calculation unit 6, and the route selection unit 7 as information including an image, a prediction result in the prediction unit 5, A warning sound generating unit 92 that generates a warning sound according to the calculation result in the interference degree calculating unit 6. The display unit 91 is realized using a display such as liquid crystal, plasma, or organic EL (Electroluminescence), while the warning sound generation unit 92 is realized using a speaker or the like. In the first embodiment, a projector is installed as the display unit 91 in the upper rear part of the driver's seat. This projector has a function of performing superimposed display on the windshield of a four-wheeled vehicle.

  The storage unit 10 includes, in addition to the detection result in the sensor unit 3, the trajectory generated by the trajectory generation unit 4, the prediction result in the prediction unit 5, the interference degree calculation result in the interference degree calculation unit 6, and the course in the route selection unit 7. The selection result and the operation selected by the operation selection unit 41 of the trajectory generation unit 4 are stored. This storage unit 10 includes a ROM (Read Only Memory) in which a program for starting a predetermined OS (Operation System), a course setting program according to the first embodiment, and the like, calculation parameters and data for each process, etc. This is realized by using a RAM (Random Access Memory) for storing. The storage unit 10 can also be realized by providing an interface on which a computer-readable recording medium can be mounted on the course setting device 1 and mounting a recording medium corresponding to this interface.

  The course setting device 1 having the above-described functional configuration is an electronic device (computer) including a CPU (Central Processing Unit) having calculation and control functions. The CPU provided in the course setting device 1 performs arithmetic processing related to the course setting method according to the first embodiment by reading from the storage unit 10 various information including information stored and stored in the storage unit 10 and the above-described course setting program. Execute.

  The course setting program according to the first embodiment can be recorded on a computer-readable recording medium such as a hard disk, a flexible disk, a CD-ROM, a DVD-ROM, a flash memory, or an MO disk and can be widely distributed. It is.

  Next, the course setting method according to the first embodiment of the present invention will be described. FIG. 2 is a flowchart showing an outline of processing of the course setting method according to the first embodiment. In the following description, it is assumed that all objects to be predicted move on a two-dimensional plane. However, the course setting method according to the first embodiment applies to an object moving in a three-dimensional space. Is applicable. Further, the present invention can be applied to a case where one object has a plurality of degrees of freedom (for example, an object such as a robot arm having 6 degrees of freedom).

  First, the sensor unit 3 detects the position and internal state of an object within a predetermined range with respect to the host vehicle, and stores the detected information in the storage unit 10 (step S1). Hereinafter, it is assumed that the position of the object is the value of the center of the object, and the internal state of the object is specified by the speed (speed v, direction θ). In this step S1, the internal state of the vehicle is also detected and stored in the storage unit 10 as a matter of course.

Next, by using the detection result input by the sensor unit 3, the trajectory generation unit 4 generates a trajectory on time and space composed of time and space for each object (step S <b> 2). FIG. 3 is a flowchart showing details of the trajectory generation processing in the trajectory generation unit 4. In this figure, the total number of objects detected by the sensor unit 3 (including the own vehicle) is K, and an operation for generating a locus for one object O _k (1 ≦ k ≦ K, k is a natural number) is N. This is performed _k times (in this sense, k and N _k are both natural numbers). Further, the time for generating the locus (trajectory generation time) is T (> 0).

  In the first embodiment, by appropriately determining the trajectory generation time T (and the operation time Δt described later), it becomes possible to predict changes in the external environment such as the course of another vehicle in a practical calculation time. This can be said in common with other embodiments of the present invention.

By the way, when performing the calculation described below, the prediction calculation is terminated at the trajectory generation time T, not whether or not the vehicle has reached a preset location (destination or intermediate location similar to the destination). It is important in terms of technical idea to have a configuration. In general, there is no place on the road where safety is guaranteed in advance. For example, as shown in FIG. 4, when the prediction is made that the own vehicle O ₁ traveling on the three-lane road Rd sequentially reaches preset locations Q ₁ , Q ₂ , and Q ₃ , the set location Taking into consideration the case where the own vehicle O ₁ travels almost straight in the same lane toward the vehicle, the other vehicle O ₃ takes the route B ₂ to avoid danger by taking the route B _3. There is a risk of entering the lane in which ₁ is traveling. Thus, in the conventional route prediction calculation, it is not guaranteed in advance that the own vehicle O ₁ is safe to travel to a preset location.

In the first embodiment, since the optimal course is determined each time without determining the destination or the like to be reached by the vehicle O ₁ in advance, under the same situation as FIG. 4, for example, FIG. The route B ₁ as shown in FIG. 5 can be selected as the route of the host vehicle O ₁ , and it is possible to appropriately avoid the danger when the host vehicle O ₁ travels and to ensure safety.

  Note that, instead of the trajectory generation time T, the cutoff condition for the prediction calculation may be determined by a trajectory generation length indicating the length of the trajectory to be generated. In this case, it is preferable to adaptively change the trajectory generation length according to the speed of the object (or the speed of the host vehicle).

First, initialization is performed so that the value of the counter k for identifying an object is set to 1, and the value of the counter n indicating the number of times of locus generation for the same object is set to 1 (step S201). Note that the following description will be described as a process for the general object O _k.

Next, the trajectory generation unit 4 reads the result detected by the sensor unit 3 from the storage unit 10, and sets the read detection result as an initial state (step S202). Specifically, the time t is set to 0, and the initial position (x _k (0), y _k (0)) and the initial internal state (v _k (0), θ _k (0)) are respectively transmitted from the sensor unit 3. Input information (x _k0 , y _k0 ) and (v _k0 , θ _k0 ).

Subsequently, the operation selection unit 41 performs one operation u _k (t) to be performed during the subsequent time Δt from among a plurality of selectable operations according to the operation selection probability given in advance to each operation. Is selected (step S203). Operation u operation selection probability for selecting the _kc p (u _kc) is defined by associating the elements with a predetermined random number, for example a set of selectable operations as _{u k (t) {u kc} }. In this sense, a different operation selection probability p (u _kc ) may be given for each operation u _kc , or an equal probability may be given to all elements of the operation set {u _kc }. In the latter case, p (u _kc ) = 1 / (total number of selectable operations). The operation selection probability p (u _kc ) can be defined as a function that depends on the position and internal state of the vehicle and the surrounding road environment.

In general, the operation u _kc is composed of a plurality of elements, and the contents of selectable operations differ depending on the type of the object O _k . For example, when the object _Ok is a four-wheeled vehicle, the acceleration and angular velocity of the four-wheeled vehicle are determined by the degree of steering and the degree of depression of the accelerator. In view of this, the operation u _kc to be performed on the object O _k is a four-wheeled vehicle is determined by elements including acceleration and angular velocity. On the other hand, when the object _Ok is a person, the operation u _kc can be designated by the speed.

A more specific setting example of the operation u _kc will be given. If the object O _k is an automobile, the acceleration -10~ + 30km / h / sec, -7~ + 7deg / sec taken in the range (designated orientation in both sign) of the steering angle, the object O _k is a human In this case, the speed is set to 0 to 36 km / h and the direction is set to 0 to 360 deg. In addition, all the quantities described here are continuous quantities. In such a case, an appropriate set of discretizations may be used to make the number of elements of each operation finite, and a set of operations {u _kc } may be configured.

Thereafter, the object operation unit 42 operates the operation u _kc selected in step S203 for a time Δt (step S204). The time Δt is preferably smaller in terms of accuracy, but may be a value of about 0.1 to 0.5 (sec) in practice. In the following description, it is assumed that the trajectory generation time T is an integer multiple of Δt, but the value of T may be variable according to the speed of the object O _i , and may not be an integer multiple of Δt.

Subsequently, the determination unit 43, along with determining whether the internal state of the object O _k after operating the operation u _kc at step S204 satisfies a predetermined control condition (step S205), the operation u _kc position of the object O _k after operating determines whether the movable region (step S206). Of these, determines the control condition in step S205 is defined according to the type of object O _k, for example, when the object O _k is a four-wheeled vehicle, and the speed range after the operation of step S204, step S204 The maximum acceleration G after the operation is determined by the vehicle G or the like. On the other hand, the movable region determined in step S206 refers to a region such as a road (including a roadway and a sidewalk). Hereinafter, the case where the object is located in the movable region is expressed as “the moving condition is satisfied”.

As a result of the determination in the determination unit 43 described above, when there is a condition that does not satisfy even one (No in Step S205 or No in Step S206), the process returns to Step S202. In contrast, the determination in the determination unit 43 result, Yes, Yes, and step S206 in the case (step S205 that the position and the internal state of the object O _k after the operation u _kc terminated in step S204 satisfies all conditions ), The time is advanced by Δt (t ← t + Δt), the position after the operation of step S204 is (x _k (t), y _k (t)), and the internal state is (v _k (t), θ _k ( t)) (step S207).

The processes in steps S202 to S207 described above are repeated until the trajectory generation time T is reached. That is, when the time t newly defined in step S207 has not reached T (No in step S208), the process returns to step S203 and is repeated. On the other hand, if the newly defined time t has reached T at step S207 (Yes in step S208), and outputs the trajectory relative to the object O _k, and stores it in the storage unit 10 (step S209).

Figure 6 is a time for the object _{O k t = 0, Δt,} 2Δt, ···, schematically the trajectory of the object O _k generated by repeating a series of processes ranging from step S203 to step S207 in T FIG. The trajectory P _k (m) (1 ≦ m ≦ N _k , where m is a natural number) shown in the figure is a three-dimensional space-time (x, y, t) of two-dimensional space (x, y) and one-dimensional time (t). ) If the locus P _k (m) is projected onto the xy plane, a predicted course of the object O _k in the two-dimensional space (x, y) can be obtained.

After step S209, if the value of the counter n has not reached N _k (No in step S210), the value of the counter n is incremented by 1 (step S211), the process returns to step S202, and the processing of steps S203 to S208 described above is performed. Repeat until the locus generation time T is reached.

When the counter n reaches N _k in step S210 (Yes in step S210), generation of all trajectories for the object O _k is completed. 7, one object O _k N _k pieces of trajectory P _k generated for _{(1), P k (2} ), ···, P k (N k) trajectory set consisting of {P _k ( n)} is an explanatory view schematically showing three-dimensional space-time. The starting point of each trajectory forming the elements of the trajectory set {P _k (n)}, that is, the initial position (x _k0 , y _k0 , 0) is the same (see step S202). Note that FIG. 7 is merely a schematic diagram, and the value of N _k can take a value of about several hundred to several tens of thousands, for example.

When the counter n reaches N _k in step S210, if the object identification counter k has not reached the total number K of objects (No in step S212), the counter k is incremented by 1 and the number of times of locus generation is counted. The value of n is initialized to 1 (step S213), and the process returns to step S202 to repeat the process. On the other hand, when the object counter k reaches K (Yes in step S212), the trajectory generation for all the objects is completed, so the trajectory generation process in step S2 is terminated, and the process proceeds to the subsequent step S3. .

A set of trajectories that can be taken by a plurality of objects existing within a predetermined range of the three-dimensional space-time by performing a predetermined number of times of trajectory generation processing on all objects detected by the sensor unit 3 in this way. A spatio-temporal environment consisting of FIG. 8 is an explanatory diagram schematically illustrating a configuration example of a spatiotemporal environment. Space environment Env when shown in FIG. (P _1, P _2), the trajectory set of objects _{O 1 {P 1 (n 1} )} ( indicated by a solid line in FIG. 8) and the trajectory set of the object O ₂ {P ₂ ( n ₂ )} (indicated by a broken line in FIG. 8). More specifically, in the spatiotemporal environment Env (P ₁ , P ₂ ), two objects O ₁ and O ₂ move along a flat and straight road R such as an expressway toward the + y-axis direction. It shows the spatio-temporal environment. In the first embodiment, since the trajectory generation is performed independently for each object without considering the correlation between the objects, the trajectories of different objects may intersect in time and space.

8, the density per unit volume of the trajectory set in each region _{{P k (n k)}} (k = 1,2) of the time-space, the density of existence probability of the object O _k in each area of the space-time ( Hereinafter, this is referred to as “space-time probability density”. Therefore, by using the space environment Env (P _1, P ₂₎ when configured by the trajectory generation processing at step S2, it is possible to determine the probability of passing a predetermined region on when the object O _k is a three-dimensional space It becomes. Note that the spatiotemporal probability density described above is merely a probability concept in spatiotemporal space, and is not necessarily 1 when the sum of the values of one object in spatiotemporal space is taken.

By the way, when a specific value of the trajectory generation time T is set in advance as a fixed value, the probability density distribution in space-time becomes uniform when the trajectory is generated beyond the value T. It is preferable to set a value that has no meaning even if it is calculated. For example, if the object is a four-wheeled vehicle and the four-wheeled vehicle is traveling normally, the maximum value may be about T = 5 (sec). In this case, if the operation time Δt in step S204 is about 0.1 to 0.5 (sec), a series of processing from step S203 to step S207 is performed to generate one trajectory P _k (m). Repeat 10 to 50 times.

  It should be noted that a method for reading different types of roads currently being traveled from map data using position data by setting different trajectory generation times T for different types of roads such as expressways, ordinary roads, and two-lane roads, image recognition, etc. It is preferable to perform switching by a method of reading the type of road by a road recognition device to which is applied.

  Also, the probability density distribution in space-time is statistically evaluated by using the trajectory calculated up to the trajectory generation time T. When the distribution is constant, the trajectory generation time T is reduced and the distribution becomes constant. If not, it is also preferable to perform adaptive control that increases the generation time.

  Furthermore, it is also possible to prepare a plurality of routes that the host vehicle can take in advance and predict the trajectory generation time T as a time when the probability that the route of the host vehicle and the path of another object intersect is constant. It is. In this case, when the predicted time is increased by Δt, the termination condition may be set when the risk increment for each course that the vehicle can take is constant. In such a configuration, in order to obtain a material for determining the route to be taken at present in order to ensure safety, it goes without saying that the endpoints on the future side of the route that the vehicle can take are set to be widely distributed spatially. Nor.

After the trajectory generation processing for each object described above, the prediction unit 5 performs probabilistic prediction of the course that each object can take (step S3). Hereinafter, as a specific prediction calculation process in the prediction unit 5, a probability that a specific trajectory P _k (m) is selected in the trajectory set {P _k (n _k )} generated for the object O _k is shown. Although the case where it calculates | requires is demonstrated, of course, this prediction calculation is only an example.

と求められる。したがって、物体Ｏ_kにＮ_k本の軌跡集合｛Ｐ_k（ｎ_k）｝が与えられたとき、物体Ｏ_kが取りうる一つの軌跡Ｐ_k（ｍ）が選ばれる確率ｐ（Ｐ_k（ｍ））は、

となる。 When N _k trajectories of the object O _k are generated, the probability p (P _k (m)) that one of the trajectories P _k (m) is an actual trajectory is calculated as follows. First, an operation sequence {u _km (t)} for realizing the trajectory P _k (m) of the object O _k is {u _km (0), u _km (Δt), u _km (2Δt),. If u _km (T)}, since the probability that the operation u _km (t) is selected at time t is p (u _km (t)), the operation sequence {u _km at t = 0 to T. The probability that (t)} is executed is

Is required. Accordingly, an object O when the trajectory set of N _k the {P _k (n _k)} is given to _k, the probability one trajectory P _k of the object O _k can take (m) is selected p (P _k (m ))

It becomes.

となる。ここで、ｓはｔ＝０からＴまでの操作時間Δｔの総数すなわち操作回数である。したがって、物体Ｏ_kが取りうるＮ_k本の軌跡に含まれる軌跡Ｐ_k（ｍ）の確率の総和はＮ_kｐ₀ ^sとなり、そのうちの１本の軌跡Ｐ_k（ｍ）が選ばれる確率ｐ（Ｐ_k（ｍ））は、式（３）を式（２）に代入することによって、

となる。すなわち、確率ｐ（Ｐ_k（ｍ））は軌跡Ｐ_k（ｍ）に依存しない。 Here, when all operations u _km (t) are selected with equal probability p ₀ (where 0 <p ₀ <1), Equation (1) becomes

It becomes. Here, s is the total number of operation times Δt from t = 0 to T, that is, the number of operations. Therefore, the sum of the probabilities of the trajectories P _k (m) included in the N _k trajectories that can be taken by the object O _k is N _k p ₀ ^s , and the probability p that one of the trajectories P _k (m) is selected. (P _k (m)) is obtained by substituting equation (3) into equation (2):

It becomes. That is, the probability p (P _k (m)) does not depend on the trajectory P _k (m).

In Equation (4), if the number of loci generated for all objects is the same (N), N ₁ = N ₂ =... = N _K = N (constant), so p _{(P k (m)) =} 1 / N , and becomes constant regardless of the object O _k. In this case, the prediction calculation in the prediction unit 5 is simplified by normalizing the value of the probability p (P _k (m)) to 1, and the predetermined prediction calculation can be executed more quickly.

In the prediction unit 5, based on the probability p (P _k (m)) calculated for each object O _k (k = 1, 2,..., K), per unit volume in each region of the three-dimensional space-time. The existence probability of the object Ok is _obtained . This existence probability corresponds to the spatio-temporal probability density on the three-dimensional space-time of the trajectory set {P _k (n _k )}, and the existence probability is generally large in the region where the density of the trajectory passing therethrough is high.

In subsequent step S4, the interference degree calculation unit 6 calculates the interference degree between the host vehicle and the other vehicle (step S4). FIG. 9 is a flowchart showing details of the interference degree calculation processing. In the following description, the object O ₁ is the own vehicle. For convenience of explanation, other objects _{O k (k = 2,3, ···} , K) also all assumed to be four-wheeled vehicle, referred to as other vehicle O _k. The interference degree calculation process shown in FIG. 9 includes four loop processes, and the other vehicle O is applied to all elements of the trajectory set {P ₁ (n ₁ )} of the own vehicle O ₁ obtained in step S3. _The degree of interference with all the _k trajectory sets {P _k (n _k )} is calculated individually.

The input interference level calculation unit 6 receives in step S4, the trajectory set of subject vehicle _{O 1 {P 1 (n 1} )}, all trajectories set of the other vehicle _{O k {P k (n k} )}, and the subject vehicle O ₁ and the interference level evaluation function for evaluating the interference of the other vehicle O _k. In the first embodiment, the interference degree calculation unit 6 is described as including an interference degree evaluation function. However, the interference degree evaluation function may be input from the outside. Further, it may be configured to adaptively vary the speed of the interference evaluation function road type and vehicle O _1.

First, the iterative process (Loop 1) for all the trajectories of the host vehicle O ₁ is started (step S401). At this time, _one trajectory of the trajectory set {P ₁ (n ₁ )} is selected, and subsequent processing is executed on the selected trajectory.

Then, to start the iterative process (Loop2) for the other vehicle O _k (step S402). In Loop 2, the counter for identifying other vehicles is initialized to k = 2, and the value of k is increased each time the repetition process is completed.

Among Loop2, to the other vehicle O _k, repeated processing (Loop3) is performed for all the elements of the trajectory set generated in step _{S3 {P k (n k)} } ( step S403). In this iterative process, the degree of interference defined by the repeat i.e. identifying counter n ₁ a trajectory generated for subject vehicle O ₁ of the Loop1 and the counter k for other vehicle identification r ₁ (n _1, k) And r ₁ (n ₁ , k) is set to 0 (step S404).

Then, to start the trajectory P ₁ of the vehicle O ₁ (n ₁₎ and repeating the process for evaluating the interference between the trajectory P _k of the other vehicle _{O k (n k) (Loop4} ) ( step S405). In Loop 4, the distances at the same time between the two trajectories P ₁ (n ₁ ) and the trajectory P _k (n _k ) are sequentially obtained at times t = 0, Δt,. Since the position of each trajectory in the two-dimensional space is defined as the center of each vehicle, the spatial distance between the two trajectories is smaller than a predetermined value (for example, the standard width or length of the vehicle). If the own vehicle O ₁ and the other vehicle O _k can be regarded as a collision. In this sense, it may be determined that two objects have collided even if the coordinate values of the two vehicles do not match. Hereinafter, the maximum distance that can be regarded as a collision of two vehicles (a spatial distance that interferes with each other) is referred to as an interference distance.

Figure 10 is a diagram schematically showing a relationship in the space when the trajectory P ₁ of the own vehicle O ₁ and (n ₁₎ and the trajectory P _k of the other vehicle O _k (n _k). In the case shown in the figure, the locus P ₁ (n ₁ ) and the locus P _k (n _k ) intersect at two points C ₁ and C ₂ . Therefore, in the vicinity of the two points C ₁ and C ₂ , there are regions A ₁ and A ₂ in which the distance between the two loci at the same time is smaller than the interference distance. That is, at the time when the two trajectories P ₁ (n ₁ ) and trajectory P _k (n _k ) are included in the areas A ₁ and A ₂ , it is determined that the own vehicle O ₁ and the other vehicle O _k collide. Made. In other words, time t = 0, Δt, ···, among T, then the number of passes through the area A ₁ and A ₂ is the number collision between the vehicle O ₁ and the other vehicle O _k.

  As can be seen from FIG. 10, in the spatiotemporal environment formed in the first embodiment, even if two trajectories collide once, subsequent trajectories are generated. This is because the trajectory for each object is generated independently.

とする（ステップＳ４０７）。ここで、第２項目ｃ_1k・ｐ（Ｐ_k（ｎ_k））・Ｆ（ｔ）について説明する。係数ｃ_1kは正の定数であり、例えばｃ_1k＝１とおくことができる。また、ｐ（Ｐ_k（ｎ_k））は式（２）で定義される量であり、他車Ｏ_kで１本の軌跡Ｐ_k（ｎ_k）が選ばれる確率である。最後のＦ（ｔ）は、１回の衝突における物体間の干渉の時間依存性を与える量である。したがって、物体間の干渉に時間依存性を持たせない場合には、Ｆ（ｔ）の値を一定とすればよい。これに対して、物体間の干渉に時間依存性を持たせる場合には、例えば図１１に示すように、時間が経過するとともに値が徐々に小さくなっていくような関数としてＦ（ｔ）を定義してもよい。図１１に示すＦ（ｔ）は、より直近の衝突を重要視する場合に適用される。 Result of obtaining distance of the vehicle O ₁ and the other vehicle O _k, if the subject vehicle O ₁ and the other vehicle O _k in the sense described above is determined to have collided (Yes at step S406), the degree of interference r ₁ The value of (n ₁ , k) is

(Step S407). Here, the second item c _1k · p (P _k (n _k )) · F (t) will be described. The coefficient c _1k is a positive constant. For example, c _1k = 1 can be set. Further, p (P _k (n _k )) is an amount defined by the equation (2), and is a probability that one trajectory P _k (n _k ) is selected in the other vehicle O _k . The last F (t) is an amount that gives time dependency of interference between objects in one collision. Therefore, when the time dependency is not given to the interference between objects, the value of F (t) may be constant. On the other hand, when the interference between objects is time-dependent, for example, as shown in FIG. 11, F (t) is expressed as a function that gradually decreases with time. It may be defined. F (t) shown in FIG. 11 is applied when the most recent collision is regarded as important.

If the time t has not reached T after step S407, Loop4 is repeated (No in step S408). In this case, the value of t is increased by Δt (step S409), the process returns to step S405, and Loop 4 is repeated. On the other hand, if the time t has reached T after step S407, Loop4 is terminated (Yes in step S408). In the case where the own vehicle O ₁ and the other vehicle O _k at time t that do not collide, the process proceeds directly to the determination processing whether to repeat the Loop 4 (step S408).

Through the loop 4 iteration process described above, the value of the interference degree r ₁ (n ₁ , k) increases as the number of collisions increases. After this Loop4 is completed, in Step S410, it is determined whether or not Loop3 is repeated. That is, if there is one interference evaluation with one trajectory P ₁ of the vehicle O ₁ (n ₁₎ is not performed among the generated trajectory for the other vehicle O _k (No at step S410), n _k Is set to n _k +1 (step S411), the process returns to step S403, and Loop 3 is repeated.

The other hand, when conducted interference evaluation with one trajectory P ₁ of the vehicle O ₁ (n ₁₎ all of the generated trajectory for the other vehicle O _k (Yes in step S410), trajectory P ₁ of the vehicle O ₁ (n ₁₎ and the final degree of interference r ₁ for evaluating the interference between total trajectory of the other vehicle O _k (n _{1, k)} grant (step S412), the grant The obtained value is output and stored in the storage unit 10 (step S413).

Step S413 the final degree of interference r ₁ output in (n _1, k) values of a total of trajectories of the other vehicle O _k, the probability one trajectory P _k (n _k) is selected p (P _k (N _k )). Therefore, in the equation (5), a constant regardless of the coefficient c _1k in k (e.g. c _1k = 1), F (t) is constant (e.g., 1) Distant, trajectory P ₁ of the vehicle O ₁ ( If the number of collisions between n ₁ ) and the trajectory P _k (n _k ) of the other vehicle O _k is M _1k (n ₁ , n _k ), the value of the interference degree r ₁ (n ₁ , k) is the trajectory P _k. probability _{p (P k (n k)} ) for each (n _k) those taking the sum the elements of _{M 1k (n 1, n k} ) all trajectories assemble a value obtained by multiplying {P _k (n _k)} become. This sum is none other than the probability of collision between one trajectory P ₁ (n ₁ ) of the host vehicle O ₁ and a trajectory that the other vehicle _Ok can take. Therefore, value finally obtained as an interference degree r ₁ (n _1, k) increases in proportion with the probability of collision with another car O _k one trajectory P ₁ of the vehicle O ₁ (n ₁₎ .

Subsequent to step S413, a determination process of whether or not to repeat Loop2 is performed. If (No at step S414) of the other vehicle O _k should perform interference evaluation with subject vehicle O ₁ remains the value of k is increased by one (step S415), and repeats the Loop2 returns to step S402. On the other hand, when the other vehicle O _k should perform interference evaluation with subject vehicle O ₁ is not left (Yes at step S414), the process proceeds to the next step S416.

In step S416, it is determined whether to repeat Loop1. Specifically, when the locus to be subjected to interference evaluation remains in the locus set {P ₁ (n ₁ )} of the own vehicle O ₁ (No in step S416), the value of n ₁ is increased by 1. (Step S417), the process returns to Step S401 and Loop1 is repeated. On the other hand, if there is no remaining track to be subjected to interference evaluation in the track set {P ₁ (n ₁ )} of the own vehicle O ₁ (Yes in step S416), Loop 1 is ended and the interference degree calculation process ( Step S4) ends.

  Next, a route selection process is performed according to the interference level calculated in the interference level calculation process (step S5). As described above, in the automatic driving technology of a moving body that moves over a wide range such as an automobile, at least the influence of other dynamic obstacles is not taken into account, or the influence is not required practically. In addition, there is a need for a route calculation technique that performs calculations related to avoiding collisions with dynamic obstacles in a practical time and avoids danger during traveling. In the first embodiment, two evaluation values are used in the course selection process for avoiding danger. The first evaluation value is the interference degree calculated by the interference degree calculation process, and the first route selection process is performed using this interference degree. When a plurality of courses are selected in the first course selection process, the second course selection process is performed using the second evaluation value stored in the storage unit 10. In this second course selection process, as a second evaluation value, in addition to a selection criterion that the route has a component along the route to the destination, an additional selection criterion (described later) for further narrowing the route is appropriately used. It is preferable to combine them.

  In this way, the degree of interference between the trajectory set of the other vehicle and the trajectory set of the own vehicle having different end points is evaluated, and the route selection according to the evaluated degree of interference is performed, which will be described with reference to FIG. As in the case of the vehicle, the destination, etc. that the vehicle should reach, is not determined in advance, and the optimum course is determined each time, and the danger while driving is properly avoided to ensure safety. It becomes possible to do. As a result, as in the case shown in FIG. 4, it is possible to solve a fatal problem that safety is not guaranteed even if the vehicle travels on a road toward a preset location.

FIG. 12 is a flowchart showing details of the course selection process. First, a trajectory with the minimum value of the interference degree r ₁ (n ₁ , k) calculated in the interference degree calculation process is selected (step S501).

If _one trajectory is selected as a result of selecting a trajectory with the minimum degree of interference r ₁ (n ₁ , k) (No in step S502), the position corresponding to the selected trajectory (x (t)) , Y (t)) and the operation set {u (t)} of t = 0 to T are read from the storage unit 10 and output to the operation signal transmission unit 8 (step S504). On the other hand, when a plurality of trajectories are selected as a result of selecting a trajectory having the minimum interference degree r ₁ (n ₁ , k) (Yes in step S502), the process proceeds to step S503.

  In step S503, the route has a component along the route to the destination, and the additional selection criteria stored in the storage unit 10 in advance are used, so that a plurality of trajectories selected in step S501 are selected. Among these, a trajectory that best matches the additional selection criterion is selected (step S503). As an additional selection criterion, a condition may be set such that there is almost no possibility of having duplicate values between trajectories.

Examples of additional selection criteria will be given below.
(1) Among the positions of the host vehicle O ₁ after the operation (after Δt), the position in the road width direction (x coordinate in FIG. 4 and the like) is closest to the center of the traveling lane. In this case, a trajectory that travels the most preferable position on the road is selected. When selecting a trajectory that performs the most stable positioning between t = 0 and T, the one that minimizes the sum of the positions in the road width direction after each operation at t = 0 to T is selected. It may be.

(2) Among the positions of the host vehicle O ₁ after the operation (after Δt), the position in the traveling direction (the y-coordinate positive direction in FIG. 4 and the like) is the maximum. In this case, the fastest trajectory is selected. Note that a trajectory having a maximum position in the traveling direction at t = T may be selected.

(3) The magnitude of acceleration at the initial time (t = 0) is minimum. In this case, the locus with the smoothest acceleration / deceleration is selected. When selecting the trajectory with the smoothest acceleration / deceleration in the operation sequence from t = 0 to T, select the one that minimizes the sum of the magnitudes of acceleration after each operation from t = 0 to T. It may be.

(4) The magnitude of the angular velocity at the initial time (t = 0) is minimum. In this case, a trajectory with the smoothest steering is selected. When selecting the trajectory with the smoothest steering in the operation sequence from t = 0 to T, the sum of the magnitudes of the angular velocities after each operation from t = 0 to T is the same as in (3). May be selected, and may be selected.

According to the course selection process described above, by selecting a trajectory that is most likely to avoid the danger of the host vehicle O ₁ as a trajectory in time and space, the host vehicle O ₁ eventually becomes an actual two-dimensional plane. You have chosen the course you should take above.

After only one trajectory is selected in step S503, the process proceeds to step S504 described above. It should be noted that when a plurality of trajectories are selected even with the additional selection criteria, for example, it may be set so as to automatically select the one having the minimum or maximum value of the counter n ₁ or k.

  Following the course selection process of step S5 described above, the operation signal transmission unit 8 records the history of the position (x (t), y (t)) corresponding to the locus output according to the selection result of step S5, Then, an operation signal corresponding to the operation set {u (t)} of t = 0 to T is generated and transmitted to the actuator device 11 (step S6).

  The operation signal generated and transmitted by the operation signal transmission unit 8 in step S <b> 6 varies depending on the configuration of the actuator device 11. For example, when the actuator device 11 is a mechanical device such as a steering, an accelerator, or a brake, the operation signal transmission unit 8 includes a history of the positions (x (t), y (t)) received from the route selection unit 7, Operation {u (t)} may be output as an operation signal as it is.

  On the other hand, when the actuator device 11 is a device that applies an operation torque to a mechanical device such as a steering, an accelerator, or a brake, the position (x (t), y received from the course selection unit 7 as described above). The history of (t)) and the operation set {u (t)} may be transmitted as an operation signal as they are, or the operation torque applied to these mechanical devices during operation is calculated, and the calculated result is You may send it. In the former case, the operation torque is calculated on the actuator device 11 side.

  If the actuator device 11 is a device that applies operating torque, a configuration in which the driver can switch to manual operation if a larger operating torque than the actuator device 11 is applied, that is, a configuration that can be overridden by the driver, is provided. The automatic driving system 100 can be applied as a driving operation assisting device, and an operation reflecting the driver's intention can be realized while selecting a course.

Note that the sensor unit 3 may also detect a road surface condition on which the host vehicle O ₁ travels, and perform feedback control on the actuator device 11 according to the road surface condition.

FIG. 13 is a diagram showing a display output example of a course selection result or the like on the display unit 91 of the output unit 9, and a spatiotemporal environment Env (P ₁ , P ₂₎ configured by two own vehicles O ₁ and another vehicle O ₂ . ₂ ) It is a figure which shows typically the example of a display output at the time of selecting the course in (refer FIG. 8). More specifically, in FIG. 13, the set route is superimposed and displayed on the windshield F by a translucent arrow H, and according to the degree of interference r ₁ (n ₁ , 2) between the host vehicle O ₁ and the other vehicle O _2. Among the paths that the own vehicle O ₁ can take on the two-dimensional plane, an area in which the value of the interference r ₁ (n ₁ , 2) exceeds a predetermined threshold is set as the object O ₁ (own vehicle). A case in which the windshield F is displayed in a translucent manner is illustrated. In the case shown in FIG. 13, a situation is schematically shown in which the actuator device 11 generates an operation torque that causes the steering ST to generate an operation torque that rotates the steering ST clockwise so as to proceed to a set course. .

Incidentally, in FIG. 13 have different illumination intensity in the two regions D _a and the area D _b that is displayed translucently (here, the greater in illumination regions D _a). Such illuminance difference interference degree r ₁ (n _1, 2) corresponds to the difference between the values of those who choose a course interference degree r ₁ illuminance travels to the vicinity of the area D _a (n _1, 2) Means that the value of is large. In this sense, FIG. 13, the interference degree r ₁ (n _1, 2) the value visually represent that the person who took the course can drive that avoids the danger towards the relatively small area D _b of ing.

In addition to displaying on the display unit 91, when the value of the interference degree r ₁ (n ₁ , 2) obtained corresponding to the expected course corresponding to the current operation exceeds a predetermined threshold value, The warning sound generator 92 may generate a warning sound (including sound).

The display output example in the output unit 9 is not limited to this. For example, the course and interference evaluation set by providing the function of the display unit 91 on the display screen CN (see FIG. 14) of the car navigation system. Results may be displayed. In this case, as shown in FIG. 14, the set course is displayed with an arrow H, and two regions _Da and two are displayed by adding color shading to each region on the two-dimensional plane displayed on the display screen CN. to display the difference between the interference of the D _b.

  According to the first embodiment of the present invention described above, a computer having storage means for storing at least the positions of a plurality of objects and the internal state including the speed of each object includes the position and the internal state of the plurality of objects. Is read from the storage means, and based on the read position and internal state of the object, the change of the position that each of the plurality of objects can take with the lapse of time in time and space is configured. Each of these is generated as a trajectory, and by using the generated trajectory, the paths of the plurality of objects are probabilistically predicted. Based on the predicted result, the path that the specific object can take and the other objects The degree of interference that quantitatively indicates the degree of interference with the path that can be taken is calculated, and the path that the specific object should take according to the calculated degree of interference is By selecting the among, it becomes possible to achieve a secure safety in situations that may occur as a reality.

  In addition, according to the first embodiment, an interference degree defined by using a collision probability in time and space between a specific object and another object is applied, and the spatiotemporal such that the interference degree is minimum. By selecting the trajectory above, it is possible to accurately set the path that is least likely to collide with other objects among the paths that a specific object can take.

  Furthermore, according to the first embodiment, by predicting the course of an object using a spatiotemporal environment formed on a spatiotemporal space composed of time and space, not only a static object but also a dynamic object The course can be predicted with high accuracy.

  In addition, according to the first embodiment, since the locus of the detected object is generated independently of each other, it is possible to distinguish a specific object (for example, the own vehicle) from other objects. As a result, it is possible to easily and accurately predict the risk that may occur between a specific object and other objects.

Note that the coefficient c _{1k in} equation (5) when increasing the value of the interference degree r ₁ (n ₁ , k) is not necessarily a constant. For example, the coefficient c _1k may be the magnitude of the relative speed at the time of collision between the host vehicle O ₁ and the other vehicle O _k . In general, the greater the relative speed, the greater the impact during a collision. Therefore, when the coefficient c _1k is the magnitude of the relative speed at the time of the collision between the vehicles, the impact degree of the collision between the vehicles is added to the interference degree r ₁ (n ₁ , k).

In addition, a value indicating the seriousness of damage may be substituted for the coefficient c _1k . In this case, for example, the magnitude of the relative speed between the vehicles at the time of collision is stored in the storage unit 10 in association with a damage scale evaluation value obtained by quantifying the damage scale caused by the collision or an amount of damage loss caused by the collision. In addition, the stored value may be read from the storage unit 10 and given the coefficient c _1k . When the sensor unit 3 has a function of detecting up to the object type, a damage scale evaluation value or a damage loss amount may be determined according to the object type. In this case, for example, when the colliding partner's object is a human and a vehicle, the value of the coefficient c _1k when colliding with a human is significantly larger than the value of the coefficient c _1k when colliding with another object. It is more preferable that the possibility of collision with a human being is made as low as possible by keeping it.

  By the way, the first embodiment can also be applied to a four-dimensional space-time (space three-dimensional, time one-dimensional) as described above. In this case, not only can it be applied to a vehicle traveling on a road with a height difference, but other moving objects that move in the air are also moving in the air, such as airplanes and helicopters. The present invention can also be applied when performing the course prediction.

  Here, the difference between the non-patent document 1 cited in the background art and the first embodiment will be described. Both of these two techniques predict the course of an object using the concept of probability. However, in Non-Patent Document 1, the course of an object within a predetermined range is not independently predicted. Probability calculation based on correlation is performed. For this reason, when any two of the plurality of objects collide, the course prediction of the two objects ends when the collision occurs. This means that, when considered on a three-dimensional space-time, the collision determination process after the point of intersection of two different object trajectories is not performed.

  On the other hand, in the first embodiment, since the object trajectory is generated independently for each object, even if different object trajectories intersect in the three-dimensional space-time, the collision determination process elapses for a predetermined time. Will continue until. As described above, the spatiotemporal environment generated in Non-Patent Document 1 and the spatiotemporal environment generated in the first embodiment are completely different.

であり、時空間環境を構成する物体の総数によらず、軌跡の本数の２乗のオーダ程度の計算量で済む。これに対して、非特許文献１において干渉評価を行う場合には、特定の物体（自車）とその他の物体（他車）とを区別していないため、互いの干渉評価を行う際の計算量（本実施の形態１における干渉度の算出回数に相当）が

であり、軌跡の本数のＫ乗のオーダ程度の計算量が必要になる。この結果、時空間環境を構成する物体の数が多いほど本実施の形態１の計算量との違いが顕著に大きくなっていく。 Further, in the first embodiment, since the independent route search is performed for each object without considering the correlation of the objects, the amount of calculation is less than that of Non-Patent Document 1. In particular, in the first embodiment, the number of times of calculating the interference degree for each trajectory is

Therefore, the calculation amount is about the order of the square of the number of trajectories regardless of the total number of objects constituting the spatiotemporal environment. On the other hand, when performing interference evaluation in Non-Patent Document 1, since a specific object (own vehicle) and other objects (other vehicles) are not distinguished, calculation when performing mutual interference evaluation The amount (corresponding to the number of times of calculating the interference degree in the first embodiment) is

Therefore, a calculation amount on the order of the Kth power of the number of trajectories is required. As a result, as the number of objects constituting the spatiotemporal environment increases, the difference from the calculation amount of the first embodiment becomes significantly larger.

  In addition, in Non-Patent Document 1, even when a phenomenon of a collision can be predicted, it is not possible to grasp when it occurs. This is because Non-Patent Document 1 does not seek the probability that an object will collide in the flow of time, but rather searches for the presence or absence of a collision for each state at each time. In other words, Non-Patent Document 1 does not explicitly use a spatiotemporal environment and does not reach the concept of spatiotemporal probability density.

  As described above, both of the first embodiment and Non-Patent Document 1 perform course prediction using the probability concept, and at first glance, it may give an impression as if it is a similar technique. The essence of the technical idea is completely different, and it is extremely difficult for those skilled in the art to arrive at the first embodiment from Non-Patent Document 1.

(Embodiment 2)
In the second embodiment of the present invention, the result of calculating the interference degree between the own vehicle (specific object) and the other vehicle obtained in the same manner as in the first embodiment described above is used, so that the time-space environment of the own vehicle and the surroundings is obtained. And evaluating the interference with the vehicle, and selecting a course to be taken by the vehicle based on the evaluation result. The functional configuration of the course setting device according to the second embodiment is the same as the functional configuration of the course setting device 1 according to the first embodiment (see FIG. 1). The course setting method according to the second embodiment is the same as the course setting method according to the first embodiment except for the interference degree calculation process and the course selection process.

FIG. 15 is a flowchart showing details of the interference degree calculation process (corresponding to step S4 in FIG. 2) of the course setting method according to the second embodiment. First, the iterative process (Loop 1) for all trajectories of the host vehicle O ₁ is started (step S421). At this time, _one trajectory of the trajectory set {P ₁ (n ₁ )} is selected, and subsequent processing is executed on the selected trajectory.

Then, to start the iterative process (Loop2) for the other vehicle O _k (step S422). In Loop 2, the counter for identifying other vehicles is initialized to k = 2, and the value of k is increased each time the repetition process is completed.

The iterative process (Loop 3) is performed on all the elements of the trajectory set {P _k (n _k )} generated in step S3 also for the other vehicle O _k (step S423). In this iterative process, the degree of interference determined by the repetition of Loop ₁ , that is, the counter n ₁ for identifying the trajectory generated for the own vehicle O ₁ and the counter k for identifying other vehicles, is the same as in the first embodiment. r ₁ (n ₁ , k) is set, and the value of r ₁ (n ₁ , k) is set as the trajectory generation time T (step S424).

Then, to start the trajectory P ₁ of the vehicle O ₁ (n ₁₎ and repeating the process for evaluating the interference between the trajectory P _k of the other vehicle _{O k (n k) (Loop4} ) ( step S425). In the Loop 4, two trajectories P ₁ and (n ₁₎ the distance in the same time with the trajectory P _k (n _k), the time t = 0, Δt, ···, sequentially obtained in T, the subject vehicle O ₁ It determines the presence or absence of collision of the other vehicle O _k. Defining collisions making this determination is the same as in the first embodiment, determines if the distance between the vehicle O ₁ and the other vehicle O _k is close than the interference distance and the collision.

Result of obtaining distance of the vehicle O ₁ and the other vehicle O _k, when the vehicle O ₁ and the other vehicle O _k is determined to have collided (Yes at step S426), the interference degree r ₁ (n _1, k) Is greater than the time t (time required from the initial position to the collision) (Yes in step S427), the value of r ₁ (n ₁ , k) is set to t, and t Is set to T (step S428). Therefore, in this case, Loop 4 ends (Yes in step S429).

In contrast, if subject vehicle O ₁ and the other vehicle O _k collide (Yes at step S426), the value of the interference degree r ₁ (n _1, k) is equal to or less than the time t at that time (step S427 No), the process proceeds to step S429 to determine whether or not to end Loop4. Incidentally, even when the subject vehicle O ₁ and the other vehicle O _k do not collide (No at step S426), the process proceeds to step S429.

  In step S429, if the time t has not reached T, Loop4 is repeated (No in step S429). In this case, the value of t is increased by Δt (step S430), the process returns to step S425, and Loop 4 is repeated. On the other hand, if the time t has reached T in step S429, Loop4 is terminated (Yes in step S429).

By repeating the process of Loop4 described above, the value of the interference degree r ₁ (n _1, k), of the collisions occurring between subject vehicle O ₁ and the other vehicle O _k, the time required until the collision from the initial position Is the shortest shortest collision time.

After Loop4 ends, in step S431, it is determined whether or not Loop3 is to be repeated. That is, if there is a locus that has not been evaluated for interference with one locus P ₁ (n ₁ ) of the own vehicle O ₁ among the loci generated for the other vehicle O _k (No in step S431), n _k. Is set to n _k +1 (step S432), the process returns to step S423, and Loop 3 is repeated. On the other hand, if made interference evaluation with one trajectory P ₁ of the vehicle O ₁ (n ₁₎ all of the trajectories generated for the other vehicle O _k (Yes in step S431), the other vehicle O _The interference evaluation for one locus P _k (n _k ) of _k is completed. Therefore, in this case, to impart a degree of interference r ₁ to evaluate interference between the trajectory P ₁ of the own vehicle O ₁ and (n ₁₎ and all trajectories of the other vehicle _{O k (n 1, k)} ( step S433 The output value is output and stored in the storage unit 10 (step S434).

Subsequent to step S434, a determination process is performed to determine whether or not to repeat Loop2. If (No at step S435) of the other vehicle O _k should perform interference evaluation with subject vehicle O ₁ remains the value of k is increased by one (step S436), and repeats the Loop2 returns to step S422. On the other hand, when the other vehicle O _k should perform interference evaluation with subject vehicle O ₁ is not left (Yes at step S435), the process proceeds to the next step S437.

を算出し、その算出結果を出力して記憶部１０に格納する（ステップＳ４３７）。ここでの重みα（ｋ）の値は、全て等しく定数（例えば１）としてもよいし、他車Ｏ_kの種別等の条件に応じた値を付与してもよい。この結果、自車Ｏ₁の軌跡Ｐ₁（ｎ₁）が全ての他車Ｏ₂、・・・、Ｏ_Kを含む周囲の他車全体との間での干渉を評価することが可能となる。 In the second embodiment, after the repetitive processing of Loop2 is finished, the interference level obtained by _{Loop2~Loop4 r 1 (n 1, k} ) weight according to the other vehicle O _k with respect to α (k) (> 0) and the total interference as the sum of these

, And the calculation result is output and stored in the storage unit 10 (step S437). Weight α value of (k) here, may be all equal constant (e.g. 1), may be given a value corresponding to the condition such as the type of the other vehicle O _k. As a result, it becomes possible trajectory P ₁ of the vehicle O ₁ to (n ₁₎ to evaluate interference between other vehicles entire circumference including all other vehicles O _2, · · ·, O _K .

と定義してもよい。この場合には、最も危険な物体Ｏ_kの危険度を全体干渉度として扱うことになる。この場合、次のような利点を有する。一般に、式（６）の定義にしたがう場合には、自車Ｏ₁が少数の物体と干渉するが、残り多数の物体とは干渉しないようなシーンの全体干渉度を低く計算してしまう恐れがあるため、例えば自車Ｏ₁の近傍を少数の他車が走行中であって人間の直観では非常に危険に思える状況であっても、直観に反して安全と判断されて可能性がある。これに対して、式（７）のような定義に基づいた干渉評価を行えば、自車Ｏ₁の近傍に存在する物体（他車を含む）の数に依存しない全体干渉度を与えるため、人間の直観に反して安全と判断してしまう可能性を低減することができる。 Note that the total interference R ₁ (n ₁ ) is

May be defined. In this case, the handle risk of the most dangerous object O _k overall interference level. This case has the following advantages. In general, when the definition of Equation (6) is followed, the vehicle O ₁ may interfere with a small number of objects, but the overall interference degree of a scene that does not interfere with the remaining many objects may be calculated low. Therefore, for example, even if a small number of other vehicles are running in the vicinity of the own vehicle O ₁ and seem to be very dangerous to human intuition, it may be judged safe against intuition. On the other hand, if the interference evaluation based on the definition like Expression (7) is performed, the overall interference degree independent of the number of objects (including other vehicles) existing in the vicinity of the own vehicle O ₁ is given. It is possible to reduce the possibility of being judged safe against human intuition.

In the subsequent step S438, a determination process is performed as to whether or not Loop1 is repeated. That is, when there is a remaining locus for interference evaluation in the locus set {P ₁ (n ₁ )} of the own vehicle O ₁ (No in step S438), the value of n ₁ is incremented by 1 (step S439), returning to step S421, repeating Loop1. On the other hand, if there is no remaining locus to be subjected to the interference evaluation in the locus set {P ₁ (n ₁ )} of the own vehicle O ₁ (Yes in step S438), Loop 1 is ended and the interference degree calculation process ( Step S4) ends.

FIG. 16 is a diagram schematically showing a configuration of a spatiotemporal environment to which the course setting method according to the third embodiment is applied. The spatiotemporal environment Env (P ₁ , P ₂ , P ₃ ) shown in the figure shows a case where two other vehicles exist within a predetermined range with respect to the own vehicle O ₁ . together they show _one trajectory P ₁ to (n ₁₎ in the solid line shows the trajectory P ₂ of the other vehicle O ₂ a (n ₂₎ shown by the broken line, the trajectory P ₃ of the other vehicle O ₃ and (n ₃₎ with a thick line . At this time, the interference degree r ₁ with another car O ₂ (n _{1, 2)} and interference degree r ₁ (n _1, 3) with another car O ₃ instead of treating separately, the space-time environment Env (P ₁ , P ₂ , P ₃ ) and the interference evaluation using the total interference degree R ₁ (n ₁ ), the danger of the host vehicle O ₁ can be avoided according to the surrounding environment.

Next, a course selection process (corresponding to step S5 in FIG. 2) will be described. FIG. 17 is a flowchart showing details of the course selection process. First, a trajectory having the maximum value of the total interference degree R ₁ (n ₁ ) calculated in the interference degree calculation process is selected (step S511).

  If one trajectory is selected as a result of selecting the trajectory with the maximum degree of interference (No in step S512), the history of the position (x (t), y (t)) corresponding to the selected trajectory. And the operation sequence {u (t)} are read from the storage unit 10 and output to the operation signal transmission unit 8 (step S514). On the other hand, if a plurality of trajectories are selected as a result of selecting the trajectory having the maximum degree of interference (Yes in step S512), the process proceeds to step S513.

  In step S513, by using the additional selection criterion set in advance and stored in the storage unit 10, a trajectory that best matches the additional selection criterion is selected from the plurality of trajectories selected in step S501 (step S513). ). As the additional selection criterion, the same conditions as in the first embodiment may be applied.

After only one trajectory is selected in step S513, the process proceeds to step S514 described above. In addition, when a plurality of trajectories are selected even by the additional selection criteria, for example, it is only necessary to automatically select the one having the minimum or maximum value of the counter n ₁ or k.

According to the course selection process described above, by selecting a trajectory that is most likely to avoid the danger of the host vehicle O ₁ as a trajectory in time and space, the host vehicle O ₁ eventually becomes an actual two-dimensional plane. You have chosen the course you should take above.

  The process of the operation signal transmission unit 8 following the course selection process in step S5 is the same as that in the first embodiment. The processing of the actuator device 11 that has received the operation signal from the operation signal transmission unit 8 is the same as in the first embodiment.

  According to the second embodiment of the present invention described above, a computer including storage means for storing at least the positions of a plurality of objects and the internal state including the velocity of each object is Is read from the storage means, and based on the read position and internal state of the object, the change of the position that each of the plurality of objects can take with the lapse of time in time and space is configured. Each of these is generated as a trajectory, and by using the generated trajectory, the paths of the plurality of objects are probabilistically predicted. Based on the predicted result, the path that the specific object can take and the other objects The degree of interference that quantitatively indicates the degree of interference with the path that can be taken is calculated, and the path that the specific object should take according to the calculated degree of interference is By selecting the among, as in the first embodiment, it becomes possible to achieve a secure safety in situations that may occur as a reality.

  Further, according to the second embodiment, the total interference degree obtained by summing the interference degree defined using the shortest collision time for each object and applying the sum is applied, and the spatio-temporal where the total interference degree is the maximum By selecting the trajectory above, even when the number of objects constituting the spatiotemporal environment is large, it is possible to accurately set a course with the lowest possibility of collision with other objects.

In the second embodiment, the interference degree r ₁ (n ₁ , k) may be defined in the same way as in the first embodiment, and the interference degree r ₁ (n ₁ , k) is increased by collision. As the values of the coefficient c _1k and F (t), it is possible to adopt any of various ways of definition as in the first embodiment. In this case, a trajectory that minimizes R ₁ (n ₁ ) may be selected during the course selection process.

In addition, when performing the interference evaluation in the second embodiment, the interference evaluation is performed in consideration of both the total interference degree R ₁ (n ₁ ) and the individual interference degree r ₁ (n ₁ , k). It may be.

(Other embodiments)
Up to this point, the first and second embodiments have been described in detail as the best mode for carrying out the present invention. However, the present invention should not be limited only by these two embodiments. For example, in the present invention, the operation selection unit of the trajectory generation unit may maintain the current operation only for the own vehicle. In this case, in the own vehicle, the internal state at the time of prediction is maintained, and only one operation is continuously performed. Therefore, the operation selection probability for selecting the operation is 1, and the time and space as the trajectory set of the own vehicle Only one trajectory is generated above.

FIG. 18 shows the spatiotemporal environment generated when the operation of the host vehicle is maintained as described above, and corresponds to FIG. In the spatio-temporal environment Env ′ (P ₁ , P ₂ ) shown in FIG. 18, the set of trajectories of the vehicle O ₁ in the three-dimensional spatio-temporal is composed of only one linear trajectory P ₁ (others The car O ₂ is the same as in FIG. In this way, if a model that maintains the operation of the host vehicle O ₁ is applied, it is possible to perform prediction by simplifying the situation when there are many surrounding objects, and a trajectory generation unit, prediction unit, and interference degree calculation. And the calculation amount in the route selection unit can be reduced. It should be noted that the operation selection probability of each operation that can be selected for the host vehicle O ₁ may be appropriately set by input from the input unit.

When the operation of the host vehicle O ₁ is maintained as described above, the display unit of the output unit displays the expected course of the host vehicle O ₁ , the risk level of the other vehicle traveling in a predetermined range, and the interference level. It is also possible to display according to the calculation result. In addition, a warning sound may be generated by a warning sound generation unit of the output unit.

In the course setting method according to the present invention, when performing trajectory generation processing on a space-time basis for each object, a trajectory may be generated by operating all selectable operations. An algorithm for realizing such a trajectory generation process can be realized by applying a recursive call by vertical search or horizontal search, for example. In this case, the number of elements i.e. the number of trajectories of the trajectory set to be finally generated for one object _{O k {P k (n)} } are not known until the trajectory generation process is completed for the object O _k. Therefore, when generating a trajectory that each object can take by searching all possible operations, the number of elements of the operation u _kc (t) in the operation time Δt (operation u _kc (t) is a continuous amount. In this case, a search method having an optimal calculation amount may be selected according to the degree of discretization.

  Furthermore, in the course setting method according to the present invention, an imaginary object may be arranged in addition to the actual object detected by the sensor unit, and the course of the arranged imaginary object may be predicted. More specifically, an imaginary object model that exhibits undesirable behavior for the host vehicle may be configured, and the object model may be placed at a predetermined position to predict the course. Such a fictitious object model is, for example, when a vehicle (own vehicle) traveling near an intersection where there is a shield or the like and has a poor visibility makes a course prediction, by placing it at a position where it cannot be detected from the own vehicle, It is possible to predict the danger such as a collision with an object that may jump out of the intersection. Note that the information on the imaginary object model may be stored in the storage unit in advance, and the object model may be arranged at a desired position according to the condition setting from the input unit.

  By the way, when the course setting device according to the present invention is applied in a region such as an expressway on which the traveling of only the vehicle is assumed, each vehicle is provided with a communication means for inter-vehicle communication, so that Vehicles traveling in the vicinity may exchange each other's travel status by inter-vehicle communication. In this case, the operation history of each vehicle is stored in its own storage unit, an operation selection probability for each operation is given based on the operation history, and information on the operation selection probability is also added to other vehicles. You may make it transmit. As a result, the accuracy of the route prediction is increased, and it is possible to more reliably avoid danger during traveling.

  In addition, it is possible to use GPS (Global Positioning System) as position detecting means for the present invention. In this case, the position information and movement information of the object detected by the sensor unit can be corrected by referring to the 3D map information stored in the GPS. Furthermore, it is also possible to function as a sensor unit by communicating GPS outputs with each other. In any case, highly accurate course prediction can be realized by using GPS, and the reliability of the prediction result can be further improved.

  As is clear from the above description, the present invention can include various embodiments and the like not described herein, and within the scope not departing from the technical idea specified by the claims. Various design changes and the like can be made.

It is a block diagram which shows the function structure of the automatic driving system which concerns on Embodiment 1 of this invention. It is a flowchart which shows the outline | summary of the course setting method which concerns on Embodiment 1 of this invention. It is a flowchart which shows the detail of the locus | trajectory generation process in the course setting method which concerns on Embodiment 1 of this invention. It is a figure which shows typically the locus | trajectory produced | generated in three-dimensional space time. It is a figure which shows typically the problem of the conventional course prediction calculation. It is a figure which illustrates typically the advantage of course prediction calculation in the course setting method concerning this Embodiment 1. FIG. It is a figure which shows typically the locus | trajectory set produced | generated in the three-dimensional space-time with respect to one object. It is explanatory drawing which shows the structure of a spatiotemporal environment typically. It is a flowchart which shows the detail of the interference degree calculation process in the course setting method which concerns on Embodiment 1 of this invention. It is a figure which shows typically the relationship in time space of one locus | trajectory of the own vehicle and one locus | trajectory of another vehicle. It is a figure which shows the example of the function which gives the time dependence of the interference between objects. It is a flowchart which shows the detail of the course selection process in the course setting method which concerns on Embodiment 1 of this invention. It is a figure which shows the example of a display output of the setting result in the course setting apparatus which concerns on Embodiment 1 of this invention. It is a figure which shows the display output example (2nd example) of the setting result in the course setting apparatus which concerns on Embodiment 1 of this invention. It is a flowchart which shows the detail of the interference degree calculation process in the course setting method which concerns on Embodiment 2 of this invention. It is explanatory drawing which shows another structure of a spatiotemporal environment typically. It is a flowchart which shows the detail of the course selection process in the course setting method which concerns on Embodiment 3 of this invention. It is a figure which shows typically the structure of the spatiotemporal environment formed when the model which maintains operation of the own vehicle is employ | adopted.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Course setting device 2 Input part 3 Sensor part 4 Trajectory generation part (trajectory generation means)
5 Prediction unit (prediction means)
6 Interference degree calculation unit (interference degree calculation means)
7 Course selection part (Course selection means)
8 Operation signal transmitter (operation signal transmitter)
9 Output unit (output means)
10 Storage unit (storage means)
11 Actuator device 41 Operation selection part (operation selection means)
42 Object operation part (object operation means)
43 determination unit (determination means)
91 display unit 92 a warning sound generating unit 100 automatic driving system B _1, B _2, B ₃ course CN display screen D _a, D _b regions _{Env (P 1, P 2)} , Env '(P 1, P 2), Env _{(P 1, P 2, P} 3) the space-time environment F front glass H arrow O _1, O _2, O _i object R, Rd road ST steering

Claims (22)

A path that can be taken by a specific object included in the plurality of objects and a path that can be taken by other objects by a computer having storage means that stores at least the positions of the plurality of objects and the internal state including the speed of each object. A course setting method for setting a course to be taken by the specific object by predicting
The position and the internal state of the plurality of objects are read from the storage means, and the change of the position that each of the plurality of objects can take with the passage of time is based on the read position and internal state of the object. A trajectory generation step for generating each as a trajectory in space-time composed of:
A prediction step for performing probabilistic prediction of the paths of the plurality of objects by using the trajectory generated in the trajectory generation step;
An interference degree calculating step for calculating an interference degree quantitatively indicating the degree of interference between a course that can be taken by the specific object and a course that can be taken by the other object, based on a result predicted in the prediction step;
A course selection step of selecting a course to be taken by the specific object according to the interference degree calculated in the interference degree calculation step;
The course setting method characterized by having. The degree of interference has a smaller value as the degree of interference between the path that the specific object can take and the path that the other object can take,
2. The route setting method according to claim 1, wherein the route selection step selects a route having the minimum degree of interference. The route selection step includes:
3. The route setting method according to claim 2, wherein when there are a plurality of routes having the minimum degree of interference, a route that best matches a predetermined additional selection criterion is selected from the plurality of routes. The degree of interference has a larger value as the degree of interference between the path that the specific object can take and the path that the other object can take is smaller,
2. The route setting method according to claim 1, wherein the route selection step selects a route having the maximum degree of interference. The route selection step includes:
5. The route setting method according to claim 4, wherein when there are a plurality of routes having the maximum degree of interference, a route that best matches a predetermined additional selection criterion is selected from the plurality of routes.   It further has an operation signal transmission step of generating an operation signal corresponding to a history of the position of the route selected in the route selection step and an operation sequence realizing the route, and transmitting the generated operation signal to the outside. The course setting method according to any one of claims 2 to 5. The interference degree calculating step includes:
The interference between the specific object and each of the other objects according to the number of times closer to the interference distance, which is a spatial distance at which the specific object and each of the other objects interfere with each other The course setting method according to any one of claims 1 to 6, wherein the degree value is increased or decreased by a predetermined amount. The interference degree calculating step includes:
The course setting method according to any one of claims 1 to 7, wherein a sum is obtained by weighting each interference value between the specific object and the other object. The trajectory generation step includes
An operation selection step of selecting an operation for the object from a plurality of operations;
An object operation step for operating the operation selected in the operation selection step for a predetermined time;
Determining whether or not the position and internal state of the object after operating the selected operation in the object operation step satisfy a control condition related to the control of the object and a movement condition related to the movable region of the object Steps,
Including
The course setting method according to any one of claims 1 to 8, wherein a series of processing from the operation selection step to the determination step is repeatedly performed until a trajectory generation time for generating a trajectory is reached.   The route setting method according to claim 1, further comprising an output step of outputting information about the route selected in the route selection step. A course setting device that sets a course to be taken by the specific object by predicting a course that can be taken by a specific object included in a plurality of objects and a course that can be taken by another object,
Storage means for storing at least the positions of the plurality of objects and an internal state including the speed of each object;
The position and the internal state of the plurality of objects are read from the storage means, and the change of the position that each of the plurality of objects can take with the passage of time is based on the read position and internal state of the object. Trajectory generating means for generating each as a trajectory in space-time composed of:
Prediction means for performing probabilistic prediction of the paths of the plurality of objects by using the trajectory generated by the trajectory generation means;
An interference degree calculating means for calculating an interference degree quantitatively indicating a degree of interference between a course that can be taken by the specific object and a course that can be taken by the other object, based on a result predicted by the prediction means;
A route selection means for selecting a route to be taken by the specific object according to the interference degree calculated by the interference degree calculation means;
A course setting device comprising: The degree of interference has a smaller value as the degree of interference between the path that the specific object can take and the path that the other object can take,
12. The route setting apparatus according to claim 11, wherein the route selection means selects a route having the minimum degree of interference. The route selection means includes
13. The route setting apparatus according to claim 12, wherein when there are a plurality of routes having the minimum degree of interference, a route that best matches a predetermined additional selection criterion is selected from the plurality of routes. The degree of interference has a larger value as the degree of interference between the path that the specific object can take and the path that the other object can take is smaller,
12. The route setting apparatus according to claim 11, wherein the route selection means selects a route having the maximum degree of interference. The route selection means includes
15. The route setting apparatus according to claim 14, wherein when there are a plurality of routes having the maximum degree of interference, a route that best matches a predetermined additional selection criterion is selected from the plurality of routes.   The apparatus further comprises operation signal transmitting means for generating an operation signal corresponding to a history of the position of the route selected by the route selecting means and an operation sequence for realizing the route, and transmitting the generated operation signal to the outside. The course setting device according to any one of claims 12 to 15. The interference degree calculating means includes:
The interference between the specific object and each of the other objects according to the number of times closer to the interference distance, which is a spatial distance at which the specific object and each of the other objects interfere with each other The course setting device according to any one of claims 11 to 16, wherein the degree value is increased or decreased by a predetermined amount. The interference degree calculating means includes:
The route setting device according to any one of claims 11 to 17, wherein a sum is obtained by weighting each interference value between the specific object and the other object. The trajectory generating means includes
An operation selecting means for selecting an operation for the object from a plurality of operations;
Object operating means for operating the operation selected by the operation selecting means for a predetermined time;
Determining whether or not the position and internal state of the object after the selected operation is operated by the object operation means satisfy a control condition related to the control of the object and a movement condition related to the movable area of the object Means,
Including
The series of processes from the operation selection process by the operation selection unit to the determination process by the determination unit are repeatedly performed until a trajectory generation time for generating a trajectory is reached. Course setting device.   The route setting device according to any one of claims 11 to 19, further comprising output means for outputting information related to a route selected by the route selection means.   A route setting program that causes the computer to execute the route setting method according to any one of claims 1 to 10. An automatic driving system that is mounted on a vehicle and automatically drives the vehicle,
The route setting device according to any one of claims 11 to 20,
An actuator device for operating the vehicle in response to an operation signal for realizing a route selected by the route selection means included in the route setting device;
An automatic driving system characterized by comprising:

Images