Control device, control method, and program

Abstract

A control device that controls a moving body that autonomously moves in a region where another pedestrian walks while at least temporarily performing preceding guidance on a preceding guidance target person, the control device being provided with: a recognition unit that recognizes an object including the preceding guidance target person and the other pedestrian; a prediction unit that predicts a future position of the identified other pedestrian; a path generation unit that generates a path through which the moving body should advance; and a drive control unit that controls a drive device attached to the moving body so that the moving body moves along the route, the route generation unit sets a plurality of route candidates, and generates a plurality of route candidates for each of the route candidates. Calculating a first score based on a positional relationship between the moving body and a destination point given in advance, and a second score based on a positional relationship between the preceding person to be guided and the other pedestrian from a current time point to a future time point; and selecting one path candidate and setting the selected path candidate as the path.

Description

TECHNICAL FIELD

[0001] The present application relates to a control device, a control method, and a program. BACKGROUND

[0002] In recent years, a mobile body (referred to as a robot, a micro mobile body, or the like) that autonomously moves while following a user for the purpose of transporting a user's baggage or the like is being put to practical use. An invention of a travel control device related to a micro mobile body is disclosed (Patent Literature 1). In addition, a mobile body that not only follows a user but also autonomously moves while leading a user in advance is also being researched.

[0003] PRIOR ART DOCUMENTS

[0004] PATENT LITERATURE

[0005] Patent Literature 1: Japanese Patent Application Publication No. 2023-34583

[0006] NON-PATENT LITERATURE

[0007] Non-Patent Literature 1: "Social Force Model for Pedestrian Dynamics" D. Helbing and P. Molnar, Physica A: Statistical Mechanics and its Applications, 20 May 1998.

[0008] Non-Patent Literature 2: "Human Trajectory Forecasting in Crowds: A Deep Learning Perspective", Parth Kothari, Sven Kreiss, and Alexandre Alahi, 11 January 2021 SUMMARY

[0009] PROBLEMS TO BE SOLVED BY THE INVENTION

[0010] In the conventional technology, there is a case where a path cannot be appropriately generated in accordance with a situation in which a mobile body is located. For example, consideration is not made as to how to lead a subject of leading in advance in terms of behavior.

[0011] The present application is completed in consideration of such a situation, and one of the objects is to provide a control device, a control method, and a program capable of generating a path of a mobile body in a manner that a subject of leading in advance is led in an appropriate behavior.

[0012] SOLUTION TO THE PROBLEM

[0013] The control device, the control method, and the program of the present application adopt the following structure.

[0014] (1): The control device of one aspect of the present application is a control device that controls a mobile body that autonomously moves in a region where other pedestrians are walking while at least temporarily conducting prior guidance on a prior guidance target, wherein the control device includes: a recognition unit that recognizes objects including the prior guidance target and the other pedestrians; a prediction unit that predicts future positions of the recognized other pedestrians; a path generation unit that generates a path along which the mobile body should advance in the future; and a drive control unit that controls a drive device mounted on the mobile body in such a manner that the mobile body moves along the path, the path generation unit sets a plurality of path candidates, calculates, for each path candidate, a first score based on a positional relationship between the mobile body and a destination point that is assigned in advance and a second score based on a positional relationship between the prior guidance target and the other pedestrians from the present time to a future time point, and selects one of the path candidates based on the first score and the second score and sets the path.

[0015] (2): In the aspect of the above (1), the first score is a score that becomes a more positive value as the mobile body is closer to the destination point that is assigned in advance, and the second score is a score that becomes a more positive value as the interval between the prior guidance target and the other pedestrians is maintained.

[0016] (3): In the aspect of the above (1), the future time point is a time point at which the mobile body reaches a terminal of each of the plurality of path candidates.

[0017] (4): In the aspect of the above (1), the path generation unit selects one of the path candidates based on a comprehensive score obtained by multiplying the first score and the second score by a weight value, respectively.

[0018] (5): In the aspect of the above (4), the path generation unit can generate a path that does not take into account the positional relationship between the prior guidance target and the other pedestrians by setting the weight value multiplied by the second score to zero.

[0019] (6): In the aspect of the above (4), the path generation unit changes the weight value based on an index value that indicates a degree of influence of the other pedestrians on the prior guidance target.

[0020] (7): In the aspect of the above (4), the path generation unit changes the weight value based on the number of the other pedestrians.

[0021] (8): In the aspect of the above (4), the path generation unit changes the weight value based on a distance between the mobile body and the prior guidance target.

[0022] (9): The path generating section changes the weight value based on the speed of the preceding guided object, in addition to the scheme of (4).

[0023] (10): The path generating section selects one of the path candidates based on the degree of deviation of a predicted path of the preceding guided object from an ideal path of the preceding guided object, in addition to the scheme of (1), the predicted path being derived based on the path candidate of the moving body, and the ideal path being derived based on the path candidate of the moving body.

[0024] (11): The ideal path is a path of the preceding guided object derived assuming that there are no other pedestrians in the vicinity of the moving body and the preceding guided object, in addition to the scheme of (10).

[0025] (12): The predicted path is a path of the preceding guided object derived taking into account the influence of the predicted path of the other pedestrians, based on a path predicted to be moved by the other pedestrians in the case where the moving body moves along the path candidate of the moving body, in addition to the scheme of (10) or (11).

[0026] (13): The control method of another aspect of the present application is a control method executed by a processor of a control device that controls a moving body that moves autonomously in a region where other pedestrians walk while at least temporarily performing preceding guidance on a preceding guided object, the control method including processes of recognizing objects including the preceding guided object and the other pedestrians, predicting future positions of the recognized other pedestrians, generating a path along which the moving body should advance in the future, and controlling a drive device mounted on the moving body so that the moving body moves along the path, the process of generating the path including processes of setting a plurality of path candidates, calculating, for each path candidate, a first score based on a positional relationship between the moving body and a destination point assigned in advance and a second score based on a positional relationship between the preceding guided object and the other pedestrians from the present time point to a future time point, and selecting one of the path candidates based on the first score and the second score and setting the path.

[0027] (14): The program of the other aspect of the present application causes the processor of the control device of the moving body which controls at least temporarily the moving body autonomously moving in the area where other pedestrians walk while previously guiding the previously guided object to perform the following processing: recognizing objects including the previously guided object and the other pedestrians; predicting future positions of the recognized other pedestrians; generating a path through which the moving body should advance in future; and controlling a driving device mounted on the moving body in such a manner that the moving body moves along the path, wherein the processing of generating the path includes the following processing: setting a plurality of path candidates, calculating, for each path candidate, a first score based on a positional relationship between the moving body and a destination point given in advance and a second score based on a positional relationship between the previously guided object and the other pedestrians from the current to the future time point, and selecting one of the path candidates based on the first score and the second score and setting it as the path.

[0028] Effects of Invention

[0029] According to the aspects (1) to (14), the path of the moving body can be generated in a manner that the previously guided object behaves appropriately. BRIEF DESCRIPTION OF DRAWINGS

[0030] Figure 1 is a configuration diagram of the moving body.

[0031] Figure 2 is a configuration diagram of the control device.

[0032] Figure 3 is a diagram showing an example of the path candidate.

[0033] Figure 4 is a flowchart showing an example of the processing content of the prediction unit related to the pedestrian.

[0034] Figure 5 is a diagram for explaining an example of the method of generating the ideal path.

[0035] Figure 6 is a flowchart showing an example of the processing content of the prediction unit related to the user.

[0036] Figure 7 is a diagram showing an example of the characteristics of the third score.

[0037] Figure 8 is a diagram for explaining the calculation process of the fourth score.

[0038] Figure 9 is a diagram exemplifying a scenario in which the moving body suppresses the approach of the user to the pedestrian.

[0039] Figure 10is a diagram illustrating a scenario in which the mobile body suppresses the approach of the user to the pedestrian.

[0040] Figure 11 is a flowchart illustrating an example of a flow of processing performed by the weight value adjustment section. DETAILED DESCRIPTION

[0041] SUMMARY

[0042] Embodiments of a control device, a control method, and a program of the present application will be described below with reference to the drawings. The control device of the present application is a control device that controls a drive device of a mobile body to move the mobile body. The mobile body in the present application refers to a mobile body that moves autonomously while giving a prior guide to a prior guide object in an area where pedestrians walk. The area where pedestrians walk refers to a sidewalk, an open space, a floor in a building, and the like, and can include a road. In the following description, it is assumed that the mobile body does not carry a person, but this is not a limitation. The prior guide object is, for example, a pedestrian, but can be a robot, an animal (hereinafter, referred to as a user U). The mobile body moves, for example, in front of the user U who is an elderly person while going to a destination point that is given in advance, and thus acts in such a manner that other pedestrians who obstruct the movement of the user U do not excessively approach the user U (i.e., acts in such a manner that the user U is given a passage). The user U is not limited to an elderly person, and can be a person who has a tendency to have difficulty in walking, a child, a person who shops at a supermarket, a patient who moves in a hospital, a pet who walks, and the like. Note that such an action can not be performed all the time, but can be performed temporarily. For example, in a case where the mobile body travels side by side with the user or chases the user, or in a case where a predetermined condition (e.g., presence of an obstacle, congestion of a traffic condition, or the like) is detected in the direction of travel of the user, the mobile body temporarily gives a prior guide to the user by executing an algorithm of the present application.

[0043] BASIC STRUCTURE

[0044] Figure 1 is a structural diagram of the mobile body 1. The mobile body 1 is equipped with, for example, an HMI (Human machine Interface) 10, an object sensing device 20, a drive device 30, a sensor 40, and a control device 100. These structures are supported or housed by a base 5.

[0045] The HMI 10 prompts various information to the user U, and accepts an input operation by the user U. The HMI 10 includes various display devices, a speaker, a buzzer, a touch panel, a switch, a button, a close proximity wireless communication device, and the like. For example, the HMI 10 accepts a setting of a destination point.

[0046] The object recognition device 20 is a device that generates data for recognizing other pedestrians present in the surroundings of the mobile body 1 as well as the user U. The object recognition device 20 includes, for example, a camera that takes the surroundings of the mobile body 1 as a range of photography. The object recognition device 20 can also include a sensor such as a radar device, a LIDAR (Light Detection and Ranging), an ultrasonic sensor, and an object recognition device that determines an object by performing sensor fusion processing based on the outputs of these sensors.

[0047] The drive device 30 is a mechanism for moving the mobile body 1 including the base body 5 in an arbitrary direction. The drive device 30 includes, for example, a plurality of wheels, a drive motor mounted to one or more of the wheels, and a steering device mounted to one or more of the wheels. The structure of the drive device 30 is not particularly restricted and can have an arbitrary structure. The drive device 30 moves the mobile body 1 while the front surface of the base body 5 is oriented in the direction of travel of the mobile body 1 as a matter of principle.

[0048] The sensor 40 is a sensor for detecting the behavior of the mobile body 1. The sensor 40 includes, for example, a wheel speed sensor for detecting the speed of the wheels, an acceleration sensor for detecting the acceleration acting on the mobile body 1, a yaw rate sensor mounted near the center of gravity in the horizontal direction of the base body 5, a steering angle sensor for detecting the steering angle of the steered wheels, a direction sensor for detecting the orientation in the horizontal direction of the mobile body 1, and the like.

[0049] Figure 2is a configuration diagram of the control device 100. The control device 100 is provided with, for example, an identification section 110, a path generation section 120, a prediction section 130, and a drive control section 140. The path generation section 120 is provided with a path candidate setting section 122, a score calculation section 124, and a weight value adjustment section 126. These constituent elements are realized, for example, by a hardware processor such as a CPU (Central Processing Unit) executing a program (software). Part or all of these constituent elements can be realized by a hardware (including circuitry) such as an LSI (Large Scale Integration), an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), a GPU (Graphics Processing Unit), and can also be realized by a cooperative operation of software and hardware. The program can be stored in advance in a storage device (a storage device provided with a non-transitory storage medium) such as an HDD (Hard Disk Drive), a flash memory, and can also be stored in a removable storage medium (a non-transitory storage medium) such as a DVD, a CD-ROM, and installed in the storage device by mounting the storage medium in a drive device.

[0050] The identification section 110 identifies objects including other pedestrians (hereinafter, referred to simply as pedestrians P) and the user U based on data output from the object recognition device 20. In the case where the object recognition device 20 is a camera, the identification section 110 identifies pedestrians by inputting a camera image to a learned model for identifying pedestrians. The same applies to objects other than pedestrians. In addition, in order to distinguish between the pedestrians P and the user U, the identification section 110 can also hold a plurality of images of the user U taken in advance by the camera as a template in a storage section (not shown), and determine the user U by comparing the template with a camera image. Furthermore, the identification section 110 can also utilize communication directivity, and identify the position of the user U by close proximity wireless communication with a terminal device held by the user U through the HMI 10.

[0051] [Path Candidates]

[0052] The path candidate setting section 122 of the path generation section 120 generates a plurality of path candidates Pmc(i) that become candidates for the path Rm to be followed by the moving body 1 in the future. i is an identifier of the path candidate, and is set to take a value from 1 to n (n is a natural number of 2 or more). The path and the path candidate are generated, for example, as a path in which a plurality (q) of path points arranged at a prescribed distance apart are connected. The path point indicates a place where the moving body 1 should arrive at every prescribed time (every step). Figure 3is a drawing showing an example of the path candidate Rmc(i). Here, let n=5 and q=3. The path candidate setting section 122 generates the path candidates Rmc(l), Rmc(2), Rmc(4), Rmc(5) in a manner that expands to the left and right with the path candidate Rmc(3) that extends in a straight line along the front direction Dm of the mobile body 1 as the center, for example. The intervals between the path points are constant, for example, and each path candidate is set to have a length of q steps. The path candidates are not generated in a manner that reaches the destination point given to the mobile body 1, but are generated in a manner that has a length of a constant distance x q steps. The path candidate Rmc(2), for example, is generated in a manner that turns to the left by a prescribed angle in each step. In addition, the path candidate Rmc(l) is generated in a manner that turns to the left by a larger prescribed angle in each step. As for the path candidates Rmc(4), Rmc(5), these are the reverse. This rule is only an example, and the path candidates can be generated based on other rules as long as the same tendency (gradual expansion) is shown. The path candidate setting section 122 outputs the path candidates Rmc(i) to the prediction section 130.

[0053] In Figure 3 the example, the intervals between the path points are set to be a constant distance, but the intervals can also be variable. The case where the intervals are longer is the case where movement at a relatively high speed is made therebetween, and the case where the intervals are shorter is the case where movement at a relatively low speed is made therebetween. That is, the element of speed can also be included in the path Rm and the path candidates Rmc(i). The path candidate setting section 122, for example, generates path candidates to which speed is given in terms of gentleness based on the path candidates generated so that the intervals between the path points are constant.

[0054] [Prediction (Simulation)]

[0055] The prediction section 130, when given the path candidates Rmc(i), predicts the future positions of the pedestrians P and the user U for each path candidate Rmc(i) on the premise thereof. The prediction section 130 predicts the future positions of the pedestrians P and the user U by the method described in any one of Non-Patent Literatures 1 to 3, or a method that combines them, for example. In the following example, the prediction section 130 predicts the future positions of the pedestrians P and the user U by a method based on the SFM (Social Force Model) described in Non-Patent Literature 1.

[0056] Figure 4 is a flowchart showing an example of the processing content of the prediction section 130 related to the pedestrians P. First, the prediction section 130 performs target prediction for each pedestrian P (S1). The prediction section 130, for example, observes the movement trajectory of the pedestrian P in time series, and predicts the front end of the movement vector of the pedestrian P in a state where there are no other pedestrians P, the mobile body 1, the user U around as the target (destination point) of the pedestrian P.

[0057] Next, the prediction unit 130 performs the processes of S2 to S6 for each path candidate Rmc(i). The prediction unit 130 inputs the path candidate Rmc(i) of the mobile body 1 one by one (S2), and generates an ideal path Ru#(i) in the case where the user U follows the mobile body 1, in association with the path candidate Rmc(i) (S3). The ideal path Ru#(i) is a path that is expected to be moved by the user U, assuming that there are no pedestrians P in the vicinity of the mobile body 1 and the user U. Figure 5 is a diagram for explaining an example of a method of generating the ideal path Ru#(i). Here, the generation of the ideal path Ru#(5) corresponding to the path candidate Rmc(5) is exemplified. For example, the prediction unit 130 generates the ideal path Ru#(5) of the user U in such a manner that the movement vector Vu-1 of the user U from Step 0 (calculation time point) toward the position Om-0 of the mobile body 1 in Step 1, the movement vector Vu-2 of the user U from Step 1 toward the position Om-1 of the mobile body 1 in Step 2, and the movement vector Vu-3 of the user U from Step 2 toward the position Om-2 of the mobile body 1 in Step 3, that is, in such a manner that the user U follows the mobile body 1 with a delay of one step. This generation method is only an example, and other generation methods can also be employed.

[0058] Return Figure 4 , the prediction unit 130 performs the processes of S4 and S5 for each pedestrian P (k = 1 to m) by the amount of q steps. First, assuming that the mobile body 1 and the user U move according to the path candidate Rmc(i) of the mobile body 1 and the ideal path Ru#(i) of the user U, respectively, various forces acting on the pedestrian P (k) in the current step are calculated (S4). The force is not an actual force, but a hypothetical force that is assumed to be generated due to the action of the psychology of the pedestrian P (k).

[0059] Among the hypothetical forces, for example, there are (1) an influence force Fl generated due to the approach of other objects (here, the mobile body 1 and the user U), (2) a turning force F2 that performs direction correction in such a manner that the direction is corrected toward the target after the advancing path is changed mainly due to the influence force Fl, (3) an acceleration / deceleration force F3 that performs speed adjustment in such a manner that the desired speed is restored after acceleration / deceleration is performed mainly due to the influence force Fl, and the like. Fl is a translation force acting in an arbitrary direction on a two-dimensional plane, F2 is a rotation force around the pedestrian as a center, and F3 is a translation force parallel to the movement vector of the pedestrian. Fl to F3 are expressed simply as follows, for example. In addition to this, the hypothetical forces can be calculated taking into account integral elements, differential elements, and the like.

[0060] Fl = (weight) x (coefficient) x (relative velocity) / (distance to other object)

[0061] F2 = (inertia moment) x (coefficient) x (angle difference between the latest movement vector and the direction toward the target)

[0062] F3 = weight x (coefficient) x (difference between the desired speed and the latest speed)

[0063] The prediction unit 130 assumes that the hypothetical force acts for one step, and predicts the position and speed of the pedestrian P(k) one step ahead (S5). By performing this process for q steps for each pedestrian, the prediction path Rp(k,i) of each pedestrian P(k) is generated and output (S6). This prediction path Rp(k,i) is a result of predicting the future position of each pedestrian P(k) for each path candidate Rmc(i).

[0064] Next, the prediction unit 130 predicts the future position of the user U according to the flowchart of Figure 6 Figure 6 is a flowchart showing an example of the processing content of the prediction unit 130 related to the user U.

[0065] The prediction unit 130 performs the processes of Sll to S14 for each path candidate Rmc(i). The prediction unit 130 inputs the prediction path of each pedestrian corresponding to the path candidate Rmc(i) (Sll). Next, the prediction unit 130 performs the processes of S12 and S13 for q steps. The prediction unit 130 assumes that the moving body 1 and each pedestrian P(k) move according to the path candidate Rmc(i) and the prediction path Rp(k,i) of each pedestrian, and calculates various forces acting on the user U in the current step (S12). The calculation method of the various forces can be the same as the process of S4 of Figure 4 The prediction unit 130 assumes that the hypothetical force acts for one step, and predicts the position and speed of the user U one step ahead (S13). By performing this process for q steps, the prediction path Ru(i) of the user U for each path candidate Rmc(i) is generated and output (S14). This prediction path Ru(i) is a result of predicting the future position of the user U for each path candidate Rmc(i).

[0066] [Evaluation of Path Candidates]

[0067] Returning to Figure 2 ​When the prediction paths Ru(i), the ideal paths Ru#(i), and the prediction paths Rp(k,i) corresponding to i = 1 to n are acquired from the prediction unit 130 as prediction results, the score calculation unit 124 calculates a comprehensive score SC(i) of each path candidate Rmc(i). The comprehensive score SC(i) is represented by Expression (1). α and β are coefficients, SC1(i) is a first score, and SC2(i) is a second score. The comprehensive score SC(i) indicates that the smaller the value, the better the evaluation, and the coefficients α, β, and ζ, η described later are set to positive values.

[0068] SC(i) = α · SC1(i) + β · SC2(i) (1)

[0069] The first score SC1(i) is a score based on the positional relationship of the mobile body 1 and the destination point given to the mobile body 1. For example, the first score SC1(i) is calculated in such a manner that the longer the distance X1 from the terminal point (the point farthest from the mobile body 1) of the path candidate Rmc(i) to the destination point given to the mobile body 1, the larger the value, and the shorter the distance, the smaller the value. The first score SC1(i) can use the distance X1 as it is, can use a value obtained by taking the logarithm, the exponential, or the like of the distance X1, or can be an output value of a certain function in which the distance is an input value.

[0070] The second score SC2(i) is a score based on the positional relationship of the user U and the pedestrian P from the current to the future time point (the time point at which the path is terminated). The second score SC2(i) is a score in which the more the interval between the user U and the pedestrian P is maintained, the more the value becomes positive. For example, the second score SC2(i) is calculated based on Expression (2). SC3(i) is a third score, and SC4(i) is a fourth score.

[0071] SC2(i) = ζ · SC3(i) + η · SC4(i) (2)

[0072] For example, the third score SC3(i) is calculated in such a manner that the distance between the path point of the predicted path Ru(i) of the user U and the path point of the prediction path Rp(k,i) of all the pedestrians P is found for each step, and the shorter the distance (i.e., the distance when the user U and any of the pedestrians P are closest during the q steps) X2 among all the steps, the larger the value, and the longer the distance, the smaller the value. However, in the case where the distance X2 is sufficiently long (the interval between the user U and the pedestrian P is sufficiently maintained), the third score SC3(i) can show a constant value. For example, in the case where the distance X2 is equal to or greater than a boundary value of about 2 [m], the third score SC3(i) can show zero. Figure 7is a graph showing an example of the characteristics of the third score SC3(i). The third score SC3(i) is preferably calculated in a manner that rapidly increases as the distance X2 becomes smaller in order to avoid the approach of the user U to the pedestrian P. For example, a barrier function, an exponential function, or the like is preferably used, but is not limited thereto, and the same tendency can be shown.

[0073] The fourth score SC4(i) is calculated in a manner that increases as the degree of deviation of the predicted path Ru(i) from the ideal path Ru#(i) becomes larger. For example, the score calculation portion 124 squares the distance of the path points of each step of the predicted path Ru(i) and the ideal path Ru#(i), and calculates the square of the total value as the fourth score SC4(i). Figure 8 is a graph for explaining the calculation process of the fourth score SC4(i). In the graph, r1 is the distance between the path points of the first step of the predicted path Ru(i) and the ideal path Ru#(i), r2 is the distance between the path points of the second step of the predicted path Ru(i) and the ideal path Ru#(i), and r3 is the distance between the path points of the third step of the predicted path Ru(i) and the ideal path Ru#(i). The fourth score SC4(i) is calculated as √(r1 2 + r2 2 + r3 2 , for example.

[0074] The score calculation portion 124 selects the path candidate Rmc(i) having the smallest comprehensive score SC(i) after calculating the comprehensive score SC(i) of each path candidate Rmc(i), and outputs the path candidate Rmc(i) as the path Rm of the mobile body 1 to the drive control portion 140. The drive control portion 140 controls the drive device 30 mounted to the mobile body 1 in such a manner that the mobile body 1 moves along the path Rm.

[0075] By thus processing, the mobile body 1 is controlled in such a manner that the process of guiding (inducing) the user U to the destination point by going to the assigned destination point and the process of making other pedestrians P not excessively approach the user U by the presence of the mobile body 1 are performed in parallel. This is because, for example, in a scenario in which the pedestrian P is approaching the user U, the comprehensive score of the path candidate Rmc(i) in which the mobile body 1 enters between the pedestrian P and the user U to restrain the approach of the pedestrian P should be relatively high.

[0076] Figure 9 and Figure 10 is a graph illustrating a scenario in which the mobile body 1 restrains the approach of the user U to the pedestrian P. In Figure 9 and Figure 10 , the left graph shows the moving path of the pedestrian P in a case in which the path Rm in which the coefficient β is set to 0 and only the simple going to the destination point is considered is selected, and the right graph shows the moving path of the pedestrian P in a case in which the coefficient β is sufficiently larger than 0.Figure 9 In the example of the right figure, the moving body 1 is oriented to the path Rm that bulges to the right, thereby (under the influence of the force F1) causing the pedestrian P to take a path closer to the left, thereby suppressing the pedestrian P from approaching the user U. Figure 10 In the example on the right of FIG, the vehicle 1 intentionally moves at a low speed, thereby causing the pedestrian P to take a path to the right (due to the influence F1), thereby preventing the pedestrian P from approaching the user U. In this way, the control device 100 can generate a path for the vehicle 1 so that the user U behaves appropriately.

[0077] [Weight (Coefficient) Adjustment]

[0078] The coefficients α, β, ζ, and η can each be a fixed value or a value adjusted by the weight value adjustment unit 126. The weight value adjustment unit 126 changes the movement tendency of the mobile body 1 by, for example, fixing α and making β, ζ, and η variable. By increasing the coefficients β and ζ, the mobile body 1 takes a more prioritized action of suppressing the pedestrian P from approaching the user U, but conversely, the possibility of reaching the destination point being slowed down is high. Therefore, by adjusting the coefficients according to the situation of the mobile body 1 and the user U, a more appropriate path can be generated. It should be noted that, based on input to the HMI 10, the weight value adjustment unit 126 can also implement an action mode in which the coefficient β is set to zero or an action mode in which η is set to zero.

[0079] For example, the weight adjustment unit 126 may adjust the coefficient β based on an indicator indicating the degree of influence exerted by the pedestrian P on the user U, such as the aforementioned influence F1. More specifically, if multiple pedestrians P are identified, the weight adjustment unit 126 may extract the greatest influence F1 among these pedestrians P and adjust the coefficient β such that the greater the greatest influence F1, the greater the coefficient β. Alternatively, the same process may be performed based on the sum of the influences F1 of multiple pedestrians P.

[0080] The weight value adjustment unit 126 may also adjust the coefficient β based on the number m of pedestrians P. More specifically, the weight value adjustment unit 126 may adjust the coefficient β so that the larger the number m of pedestrians P, the larger the coefficient β.

[0081] Furthermore, the weight value adjustment unit 126 may adjust the coefficient β based on both the influence F1 and the number m of pedestrians P. More specifically, the weight value adjustment unit 126 may perform the same processing as described above based on a value obtained by multiplying the maximum or total value of the influence F1 by the number m of pedestrians P.

[0082] Further, the weight value adjustment section 126 can also adjust the coefficient β based on the distance Xmu between the mobile body 1 and the user U. More specifically, the weight value adjustment section 126 can also adjust the coefficient β in such a manner that the longer the distance Xmu between the mobile body 1 and the user U, the larger the coefficient β. This is because a case where the distance Xmu between the mobile body 1 and the user U is long indicates that the user U is delaying the following of the mobile body 1 for some reason, and thus it is presumed that the mobile body 1 needs to be made to approach the user U more so as to restrain the approach of the pedestrian P.

[0083] Further, the weight value adjustment section 126 can also adjust the coefficient β based on the speed Vu of the user U. More specifically, the weight value adjustment section 126 can also adjust the coefficient β in such a manner that the slower the speed Vu of the user U, the larger the coefficient β. This is because a case where the speed Vu of the user U is slow indicates that the user U is delaying the following of the mobile body 1 for some reason, and thus it is presumed that the mobile body 1 needs to be made to approach the user U more so as to restrain the approach of the pedestrian P.

[0084] Figure 11 is an example of a flowchart showing the processing performed by the weight value adjustment section 126 in a case where all of the elements of the above-described explanation are reflected. First, the weight value adjustment section 126 determines whether or not the maximum value maxFl of the influence F1 is equal to or greater than the threshold value Thl (S20). In a case where the maximum value maxFl of the influence F1 is equal to or greater than the threshold value Thl, the weight value adjustment section 126 increases the coefficient β by a first prescribed value Al (S21).

[0085] Next, the weight value adjustment section 126 determines whether or not the number m of pedestrians P is equal to or greater than a threshold value Th2 (S22). In a case where the number m of pedestrians P is equal to or greater than the threshold value Th2, the weight value adjustment section 126 increases the coefficient β by a second prescribed value A2 (S23).

[0086] Next, the weight value adjustment section 126 determines whether or not the distance Xmu between the mobile body 1 and the user U is equal to or greater than a threshold value Th3 (S24). In a case where the distance Xmu is equal to or greater than the threshold value Th3, the weight value adjustment section 126 increases the coefficient β by a third prescribed value A3 (S25).

[0087] Next, the weight value adjustment section 126 determines whether or not the speed Vu of the user U is less than a threshold value Th4 (S26). In a case where the speed Vu is less than the threshold value Th4, the weight value adjustment section 126 increases the coefficient β by a fourth prescribed value A4 (S27).

[0088] According to the above-described embodiment, it is possible to generate the path of the mobile body 1 in such a manner that the preceding guide object (the user U) is caused to behave appropriately.

[0089] In the above description, the control device 100 is provided as a control device mounted on the mobile body 1, but is not limited thereto, and can be a control device provided at a place separate from the mobile body 1 and acquiring output data of the object detection device 20 using communication and transmitting a drive instruction signal to the drive device 30, that is, a control device that remotely controls the mobile body 1.

[0090] The embodiment described above can be expressed as follows.

[0091] A control device that controls a mobile body that autonomously moves in a region where other pedestrians are walking while at least temporarily performing prior guidance on a prior guidance target,

[0092] The control device includes:

[0093] one or more storage media that hold computer-readable instructions; and

[0094] a processor connected to the one or more storage media,

[0095] The processor executes the computer-readable instructions to:

[0096] identify objects including the prior guidance target and the other pedestrians;

[0097] predict future positions of the identified other pedestrians;

[0098] generate a path along which the mobile body should advance in the future; and

[0099] control a drive device mounted on the mobile body so that the mobile body moves along the path,

[0100] The process of generating the path includes a process of setting a plurality of path candidates, calculating, for each path candidate, a first score based on a positional relationship between the mobile body and a destination point assigned in advance and a second score based on a positional relationship between the prior guidance target and the other pedestrians from a current time point to a future time point, and selecting one of the path candidates based on the first score and the second score and setting the path.

[0101] The above describes a specific embodiment of the present application using the embodiment, but the present application is not limited at all to such an embodiment, and various modifications and substitutions can be made within a range not departing from the gist of the present application.

[0102] Reference Signs

[0103] 1 moving body

[0104] 20 object detecting device

[0105] 30 drive device

[0106] 100 control device

[0107] 110 recognition section

[0108] 120 route generation section

[0109] 122 route candidate setting section

[0110] 124 score calculation section

[0111] 126 weight value adjustment section

[0112] 130 prediction section

[0113] 140 drive control section

Claims

1. A control device that controls a mobile body autonomously moving in a region where other pedestrians are walking while at least temporarily conducting prior guidance on a prior guidance target, wherein the control device comprises: a recognition unit that recognizes objects including the prior guidance target and the other pedestrians; a prediction unit that predicts a future position of the recognized other pedestrians; a path generation unit that generates a path along which the mobile body should advance in the future; and a drive control unit that controls a drive device mounted on the mobile body in such a manner that the mobile body moves along the path, the path generation unit sets a plurality of path candidates, calculates, for each path candidate, a first score based on a positional relationship of the mobile body with respect to a destination point that is assigned in advance and a second score based on a positional relationship of the prior guidance target with respect to the other pedestrians from a current time point to a future time point, selects one of the path candidates based on the first score and the second score, and sets the path.

2. The control device according to claim 1, wherein the first score is a score that becomes a positive value as the mobile body is closer to the destination point that is assigned in advance, and the second score is a score that becomes a positive value as an interval between the prior guidance target and the other pedestrians is maintained.

3. The control device according to claim 1, wherein the future time point is a time point at which the mobile body reaches a terminal of each of the plurality of path candidates.

4. The control device according to claim 1, wherein the path generation unit selects one of the path candidates based on a comprehensive score obtained by multiplying the first score and the second score by a weight value, respectively.

5. The control device according to claim 4, wherein the path generation unit is capable of generating a path that does not take into account the positional relationship of the prior guidance target with respect to the other pedestrians by setting the weight value multiplied by the second score to zero.

6. The control device according to claim 4, wherein the path generation unit changes the weight value based on an index value that indicates a degree of influence of the other pedestrians on the prior guidance target.

7. The control device according to claim 4, wherein the path generation unit changes the weight value based on a number of the other pedestrians.

8. The control device according to claim 4, wherein the path generation unit changes the weight value based on a distance of the mobile body from the prior guidance target.

9. The control device according to claim 4, wherein the path generation unit changes the weight value based on a speed of the prior guidance target.

10. The control device according to claim 1, wherein the path generation unit selects one of the path candidates further based on a degree of deviation of a predicted path of the prior guidance target from an ideal path of the prior guidance target, the predicted path being derived based on the path candidate of the mobile body, and the ideal path being derived based on the path candidate of the mobile body.

11. The control device according to claim 10, wherein ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ The ideal path is a path of movement of the preceding guide object derived on the assumption that the other pedestrians do not exist in the surroundings of the mobile body and the preceding guide object.

12. The control device according to claim 10 or 11, wherein The ideal path is a path of movement of the preceding guide object derived on the basis of a predicted path of the other pedestrians predicted on the assumption that the mobile body moves along the path candidate of the mobile body.

13. A control method of a control device that controls a mobile body that autonomously moves in an area where other pedestrians walk while at least temporarily performing preceding guidance of a preceding guide object, wherein The control method includes the following processes: identifying objects including the preceding guide object and the other pedestrians; predicting future positions of the identified other pedestrians; generating a path along which the mobile body should advance in future; and controlling a drive device mounted on the mobile body in such a manner that the mobile body moves along the path, the process of generating the path includes the following processes: setting a plurality of path candidates, for each path candidate, calculating a first score based on a positional relationship of the mobile body with a destination point assigned in advance and a second score based on a positional relationship of the preceding guide object and the other pedestrians from the current to a future time point, and selecting one of the path candidates based on the first score and the second score and setting it as the path.

14. A program that causes a processor of a control device that controls a mobile body that autonomously moves in an area where other pedestrians walk while at least temporarily performing preceding guidance of a preceding guide object to execute the following processes: identifying objects including the preceding guide object and the other pedestrians; predicting future positions of the identified other pedestrians; generating a path along which the mobile body should advance in future; and controlling a drive device mounted on the mobile body in such a manner that the mobile body moves along the path, wherein the process of generating the path includes the following processes: setting a plurality of path candidates, for each path candidate, calculating a first score based on a positional relationship of the mobile body with a destination point assigned in advance and a second score based on a positional relationship of the preceding guide object and the other pedestrians from the current to a future time point, and selecting one of the path candidates based on the first score and the second score and setting it as the path.

Images