JP2021163215A - Mobile object control device, mobile object control program and mobile object - Google Patents

Abstract

To provide a mobile object control device which can move in harmony with surrounding pedestrians.SOLUTION: A mobile object control device 6 comprises: mobile object information acquisition means 610 for acquiring mobile object information 600; pedestrian information acquisition means 611 for acquiring pedestrian information 601; evaluation value calculation means 612 for calculating a plurality of evaluation values for a plurality of movement patterns on the basis of these pieces of information; and movement control mean 613 for selecting a specific movement pattern by comparing the plurality of evaluation values with one another, and outputting a movement command by the movement pattern to a movement unit 5. The evaluation value calculation means 612 calculates the evaluation values by synthesizing first virtual attractive force acting to attract a mobile object to a target position, second virtual attractive force acting to attract the mobile object to a pedestrian existing in front of the mobile object and going toward the same direction as that of the mobile object, first virtual repulsive force acting between the pedestrian going toward a mobile object side and the mobile object, and second virtual repulsive force acting between the pedestrian going toward the mobile object side and other pedestrians.SELECTED DRAWING: Figure 2

Description

The present invention relates to a mobile control device, a mobile control program, and a mobile body.

Conventionally, various control methods for controlling a robot moving in an environment in which a pedestrian exists are known. For example, Patent Document 1 discloses a robot that predicts a pedestrian's trajectory and determines its own avoidance behavior when the robot passes by a pedestrian.

Japanese Unexamined Patent Publication No. 2012-20818

As described above, the robot disclosed in Patent Document 1 determines its own avoidance behavior on the assumption that it passes by a pedestrian approaching itself from the front. However, in the environment, there are pedestrians who walk in various directions in addition to pedestrians who come toward themselves from the front. Therefore, it is difficult to move in harmony with the surrounding pedestrians because the flow of crowds is disturbed only by deciding the avoidance behavior of oneself considering only the pedestrians who come toward oneself from the front.

The present invention has been made in view of such circumstances, and an object of the present invention is to provide a moving body control device, a moving body control program, and a moving body capable of moving in harmony with surrounding pedestrians. do.

The present invention solves the above problems, and the mobile control device according to the embodiment of the present invention is
A moving body control device that controls the moving means of the moving body so as to move the moving body to a predetermined target position in an environment where a pedestrian exists.
A moving body information acquisition means for acquiring moving body information including the target position, the position of the moving body, and a velocity vector, and
A pedestrian information acquisition means for acquiring pedestrian information including the position and speed vector of the pedestrian existing around the moving body, and
An evaluation value calculating means for calculating a plurality of evaluation values for a plurality of movement patterns for the moving means based on the moving body information and the pedestrian information.
A movement control means for selecting a specific movement pattern by comparing the plurality of evaluation values and controlling the movement means based on a movement command according to the movement pattern is provided.
The evaluation value calculation means is
A first virtual attraction that acts to attract the moving object to the target position,
A second virtual attraction that exists in front of the moving body and acts to attract the moving body to the pedestrian heading in the same direction as the moving body.
A first virtual repulsive force acting between the pedestrian and the moving body toward the moving body side,
The evaluation value is calculated by synthesizing a second virtual repulsive force acting between the pedestrian toward the moving body side and another pedestrian.

Further, the mobile control program according to the embodiment of the present invention is
It is characterized in that the computer functions as the mobile control device.

Further, the moving body according to the embodiment of the present invention is
It is characterized by including the above-mentioned mobile body control device.

According to the pedestrian control device, the pedestrian control program, and the pedestrian according to the embodiment of the present invention, the evaluation value calculating means has a first virtual attractive force that acts to attract the pedestrian to the target position, and the pedestrian. A second virtual attraction that exists in front of the moving body and acts to attract the moving body to a pedestrian heading in the same direction as the moving body, and a first acting between the pedestrian and the moving body toward the moving body side. By synthesizing the virtual repulsive force of the above and the second virtual repulsive force acting between the pedestrian toward the moving body side and another pedestrian, a plurality of evaluation values are calculated for each of a plurality of movement patterns, and movement control is performed. The means selects a specific movement pattern by comparing a plurality of evaluation values, and controls the movement means based on the movement command according to the movement pattern.

Therefore, the moving body follows the pedestrian by considering the second virtual attractive force that exists in front of the moving body and acts to attract the moving body to the pedestrian heading in the same direction as the moving body. The moving body can move in harmony with the surrounding pedestrians. Further, by considering the second virtual repulsive force acting between the pedestrian toward the moving body side and the other pedestrian, the influence of the moving body on the pedestrian is reduced, so that the moving body is in the surroundings. Can move in harmony with pedestrians.

It is a schematic diagram which shows an example of the moving body 1 (robot R) which concerns on embodiment of this invention. It is a block diagram which shows an example of the moving body 1 (robot R) which concerns on embodiment of this invention. It is a schematic diagram showing a first virtual attractive force acting between the target position P _r ^goal and the robot R. The method of setting the second virtual attraction force that exists in front of the robot R and acts between the pedestrian H and the robot R heading in the same direction as the robot R and the weighting coefficient α _{i for the second virtual attraction force.} It is a schematic diagram which shows. It is a schematic diagram which shows the setting method of the 1st virtual repulsive force acting between the pedestrian H and the robot R toward the robot R side, and the weighting coefficient β _{i with respect to the 1st virtual repulsive force.} It is a schematic diagram which shows the 2nd virtual repulsive force acting between the pedestrian H toward the robot R side and another pedestrian H. It is a flowchart which shows the operation of the mobile body control device 6 which concerns on embodiment of this invention. It is a figure which shows an example of the movement pattern Cs and the time series movement pattern _Dk. It is a figure which shows the analysis result which simulated the movement state of the robot R when the second virtual attractive force was set to "none". It is a figure which shows the analysis result which simulated the movement state of the robot R when the second virtual attractive force was set to "yes". Comparison result of comparing the movement status of the robot R between the case where the second virtual attraction is set to "None" (Fig. 9) and the case where the second virtual attraction is set to "Yes" (Fig. 10). It is a figure which shows. It is a figure which shows the analysis result which simulated the movement state of the robot R when the second virtual repulsive force was set to "none". It is a figure which shows the analysis result which simulated the movement state of the robot R when the second virtual repulsive force was set to "yes". Comparison result of comparing the movement status of the robot R between the case where the second virtual repulsive force is set to "None" (Fig. 12) and the case where the second virtual repulsive force is set to "Yes" (Fig. 13). It is a figure which shows.

Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings.

FIG. 1 is a schematic view showing an example of a moving body 1 (robot R) according to an embodiment of the present invention. FIG. 2 is a block diagram showing an example of the moving body 1 (robot R) according to the embodiment of the present invention.

The moving body 1 is configured as, for example, a self-propelled robot R, and is a device that autonomously moves in an environment in which a pedestrian H exists and moves toward a predetermined target position G set in the environment. Is. The mobile body 1 may be used for any purpose such as industrial use, commercial use, home use, medical use, disaster use, and research use. Further, the environment in which the moving body 1 is used may be indoors or outdoors.

In the present embodiment, the moving body 1 will be described as being a self-propelled robot R as shown in FIG. The robot R includes, for example, a columnar main body 2, a self-position detection unit 3 that detects the current position (self-position) and posture (traveling direction) of the self-device (robot R), and walking existing around the robot R. A pedestrian detection unit 4 that detects a person H, a moving unit 5 that moves the robot R back and forth and left and right, and a moving body control process (moving body control method) that is built in the main body 2 and controls each part of the robot R. A moving body control device 6 for performing the operation and a power source 7 for supplying power to each part of the robot R are provided.

The self-position detection unit 3 is composed of, for example, a GPS sensor or a position estimation sensor by dead reckoning (rotation angular velocity sensor of wheels 50, etc.), detects the self-position and traveling direction of the robot R, and moves the detection result. It is sent to the body control device 6. The self-position detection unit 3 may detect the self-position of the robot R by receiving position information from a position transmitter (not shown) installed at each position in the environment, or the above-mentioned plurality of self-position detection units 3. It may be a combination of the above means.

The pedestrian detection unit 4 is composed of, for example, a distance measuring sensor such as a laser range sensor or a millimeter wave sensor, and is attached to the pedestrian H at heights near the shoulder and near the ankle, respectively. The pedestrian detection unit 4 irradiates a laser beam or a millimeter wave in all directions (360 degrees), and measures the time until the reflected light or the reflected wave from the pedestrian H existing in the detection range 40 returns. By detecting the relative position (distance and direction) of the pedestrian H with respect to the robot R and observing the shape of the shoulder and the movement of the feet, the traveling direction of the pedestrian H is detected. Then, the pedestrian detection unit 4 sends the detection result of the pedestrian H to the mobile control device 6. The pedestrian detection unit 4 may be composed of a camera, an ultrasonic sensor, an infrared sensor, or the like, or may be a combination of the above-mentioned plurality of means. Further, the pedestrian detection unit 4 may also have a function of detecting another moving body 1 or an obstacle.

The moving unit 5 includes a plurality of wheels 50 and a plurality of electric motors 51 to which electric power is supplied from the power source 7 to rotationally drive the plurality of wheels 50. The moving unit 5 changes the translational motion v when the robot R moves forward or backward by controlling the rotation speed and torque of the electric motor 51 based on the movement command sent from the moving body control device 6. , The angular velocity ω when the robot R turns is changed. As a function of turning the robot R, the moving unit 5 may be provided with, for example, a difference in the number of rotations of the left and right wheels 50, or may be provided with a steering mechanism for changing the direction of the front wheels 50. ..

The moving body control device 6 does not collide with the pedestrian H based on the detection result of the own position sent from the own position detection unit 3 and the detection result of the pedestrian H sent from the pedestrian detection unit 4. as to move the robot R to the target position P _r ^goal, it controls the robot R by sending movement commands to the mobile unit 5.

As a specific configuration thereof, the mobile control device 6 is composed of an HDD, SDD, ROM, RAM and the like, and is composed of a storage unit 60 for storing various information and a processor such as a CPU. It is provided with a control unit 61 that performs various calculations, and a communication unit 62 that is a communication interface between an external device (including, for example, an infrastructure facility, a centralized management device, another mobile body, and the like).

Storage unit 60, the moving body information 600 including the position and velocity vector of the target position P _r ^goal and the robot R of the robot R, a pedestrian information 601 including the position and velocity vector of the pedestrian H, evaluation value J (details 602 stores setting information 602 including various coefficients, conditions, calculation formulas, etc. necessary for calculating (described later), and a moving body control program 603.

The target position P _r ^goal included in the mobile information 600 is, for example, those set by the administrator or the like of the external device and the robot R. Further, the position and velocity vectors of the robot R included in the moving body information 600 are accumulated for a predetermined period from a past time point to the present time, and the own position and the traveling direction detected by the own position detection unit 3 The current information is updated sequentially based on the detection result of. The velocity vector may be calculated from the amount of change in the own position detected at a predetermined time interval.

The position and speed vector of the pedestrian H included in the pedestrian information 601 is the accumulation of a predetermined period from the past time point to the present time, and the detection result of the pedestrian H detected by the pedestrian detection unit 4 The current information is updated sequentially based on. When a plurality of pedestrians H are present around the robot R, pedestrian information 601 is stored for each of the plurality of pedestrians H. Further, the velocity vector may be calculated from the amount of change in the position of the pedestrian H detected at a predetermined time interval.

Control unit 61, by executing the mobile control program 603, the moving body information obtaining means 610 for obtaining mobile body information 600 including the position and velocity vector of the target position _P ^{r goal} and the robot R of the robot R, the robot R A plurality of movement patterns for the movement unit 5 based on the pedestrian information acquisition means 611, the moving body information 600, and the pedestrian information 601 for acquiring the pedestrian information 601 including the position and speed vector of the pedestrian H existing around the vehicle. A specific movement pattern is selected from a plurality of movement patterns by comparing the evaluation value calculation means 612 for calculating a plurality of evaluation values J and the plurality of evaluation values J, and a movement command based on the movement pattern is issued. It functions as a movement control means 613 that controls the movement unit 5 based on the movement control unit 5.

The calculation formula when the evaluation value calculation means 612 calculates the evaluation value J is defined by the following formula (1).

Further, the total evaluation value Js as the sum of the evaluation values J calculated over the plurality of control steps is defined by the following equation (2).

In the above equations (1) and (2), the influence of the behavior of the robot R on the surrounding pedestrian H is measured by the evaluation value J of the model called Social Force Model (SFM), and the robot R and the pedestrian H are used. It is expressed as the force acting between (virtual attraction and virtual repulsion).

That is, the evaluation value calculation unit 612, a first virtual attractive force acting to attract the robot R to the target position ^{_{P r goal ((1),}} (2) equation the first term of) and, in front of the robot R A second virtual attraction (the second term of equations (1) and (2)) that acts to attract the robot R to the pedestrian H heading in the same direction as the robot R, and walking toward the robot R side. Between the first virtual repulsive force acting between the person H and the robot R (the third term of equations (1) and (2)) and the pedestrian H and the other pedestrian H toward the robot R side. The evaluation value J or the total evaluation value Js is calculated by synthesizing the acting second virtual repulsive force (the fourth term of the equations (1) and (2)). In the above equations (1) and (2), α, β, and γ are coefficients for adjusting the balance of each term, and are appropriately set by simulation, actual machine test, or the like. Further, in the above (2), _w ^r goal to be multiplied respectively in terms _{of ^{w r, h goal, w r}} , h int, w h, h int is the sum of each term from 0 to 1 A coefficient for normalizing to a range.

Figure 3 is a schematic diagram showing a first virtual attractive force acting between the target position P _r ^goal and the robot R. First virtual attractive force _f ^{r goal,} as shown in FIG. 3, the request speed vector of the current velocity vector _{V r} of the robot R, required for the robot R to move toward the target position _P ^{r goal} ( Target position arrival velocity vector) It is defined as a force for changing to _{V r} ^desired.

FIG. 4 shows a second virtual attractive force existing in front of the robot R and acting between the pedestrian H and the robot R heading in the same direction as the robot R, and a weighting coefficient α _{i with respect to the second virtual attractive force.} It is a schematic diagram which shows the setting method of. As shown in FIG. 4, the second virtual attractive force ^fr _{, hi} goal is a pedestrian H that exists in front of the robot R and heads in the same direction as the robot R with _{the current velocity vector Vr of the robot R.} toward the _i request speed vector robot R is required to move (pedestrian tracking speed vector) V _r, it is defined as the force for changing the _hi ^Desired.

Further, the weighting coefficient α _{i with} respect to the second virtual attractive force ^fr _{, hi} goal is set by the following equation (3).

When the evaluation value calculating means 612 synthesizes the second virtual attractive force into the evaluation value J or the comprehensive evaluation value Js by the above equations (1) and (2), the weighting coefficient α _{i is based on the above equation (3).} To set. Accordingly, the evaluation value calculating means 612 in front of the robot R, and pedestrian _{H i} (Fig. 4 pedestrian _{H 1} In corresponds.) Toward the same direction as the robot R, the target position _P ^{r goal} The second virtual attraction that acts between the _{pedestrian Hi} and the robot R becomes larger as the person walks toward the _{robot (the smaller the angle | δ i} |). It is weighted. In the example shown in FIG. 4, the weighting factor alpha ₁ for pedestrians _{H 1} is "cos | δ ₁ |" is set in the weighting factor for the pedestrian _{H 2, H 3 α 2,} α 3 are set to "0" NS.

At that time, the evaluation value calculating means 612 further includes a pedestrian _{Hi (corresponding to the pedestrian H 1 in} FIG. 4) existing in front of the robot R and heading in the same direction as the robot R, and the robot R. The second virtual attraction may be weighted in consideration of the relative speed between the two. For example, the condition of the above (3), by further adding a condition ^{_"v r} _desired -v _hi <-1", if the speed of the pedestrian _{H i} is slower than the speed of the robot R , it may be the weighting factor alpha _i for the walker H _i is set to "0".

FIG. 5 is a schematic diagram showing a method of setting a first virtual repulsive force acting between the pedestrian H and the robot R toward the robot R side and a weighting coefficient β _{i for the first virtual repulsive force.} As shown in FIG. 5, the first virtual repulsive force _{fr, high} ^int is defined as a force acting in the direction opposite to the robot R side on the straight line connecting the pedestrian H and the robot R. The size of the first virtual repulsive force f _{r, hi} ^int, for example, as a parameter the distance between the pedestrian H _i and the robot R, is calculated by substituting a predetermined calculation formula containing the exponential function.

Further, the weighting coefficient β _{i with} respect to the first virtual repulsive force _{fr, hi} ^int is set by the following equation (4).

When the evaluation value calculating means 612 synthesizes the first virtual repulsive force into the evaluation value J or the comprehensive evaluation value Js in the above equations (1) and (2), the weighting coefficient β _{i is based on the above equation (4).} To set. Therefore, in the evaluation value calculation means 612, the more the pedestrian _{Hi (corresponding to the pedestrian H 2 in} FIG. 5) toward the robot R side is walking toward the robot R (angle | θ _i |). The first virtual repulsive force acting between the pedestrian _Hi and the robot R becomes larger), and the first virtual repulsive force is weighted. In the example shown in FIG. 5, the weighting factors β ₂ and β _{3 for pedestrians H 2} and H ₃ are set to be larger than the weighting factors β ₁ for pedestrian H _1.

The weighting coefficients α _i and β _i are set by using other functions such as the sigmoid function instead of using the trigonometric function in the above equations (3) and (4). You may.

FIG. 6 is a schematic view showing a second virtual repulsive force acting between the pedestrian H toward the robot R side and another pedestrian H. Second virtual repulsive force f _{hi, hj} ^int is defined, as shown in FIG. 6, on a straight line connecting the pedestrian H _i and other pedestrian H _j, as a force acting in the opposite direction to the pedestrian side Will be done. Second virtual repulsive force f _hi, the magnitude of _hj ^int, for example, the distance between the pedestrian H _i and other pedestrian H _j as a parameter, by substituting a predetermined calculation formula containing the exponential function It is calculated. Assuming that a group is formed by a plurality of pedestrians, for example, when the pedestrian _Hi and another pedestrian H _j belong to different groups, the virtual repulsive force always acts, and the pedestrian When _Hi and other pedestrians H _j belong to the same group, a virtual repulsive force acts when they are closer than a predetermined approach distance, and a virtual attractive force acts when they are separated from a predetermined distance. good.

(Operation of robot R and mobile control device 6)
Next, the operation of the robot R and a mobile control unit 6 when moving the environment in which the robot R exists pedestrian H _i.

FIG. 7 is a flowchart showing the operation of the mobile control device 6 according to the embodiment of the present invention. FIG. 8 is a diagram showing an example of the movement pattern Cs and the time-series movement pattern _Dk.

Mobile control device 6, for example, that set ^the target position P _r ^goal moving to the target position P _r ^goal is permitted as a start condition, a series of processes (mobile control shown in the flowchart of FIG. 7 Processing) is repeatedly executed for each control step. For each control step, for example, a predetermined time interval (control cycle) is set according to the specifications of the robot R (translational motion v, maximum value of angular velocity ω, radius of detection range 40, etc.) and congestion in the environment. Will be done.

Then, the moving body control device 6 sequentially acquires the latest moving body information 600 and pedestrian information 601 by executing the moving body control process for each control step, and these moving body information 600 and pedestrian information. the movement command corresponding to 601 by sequentially outputting the moving unit 5 is moved toward the robot R to the target position P _r ^goal while avoiding a collision with a pedestrian H _i. As a result, as the end condition that the robot R reaches the target position P _r ^goal, operation of the mobile control device 6 is completed. In the case where the new target position P _r ^{goal has} been set, the above-described operation is executed again.

Further, as described above, when the mobile control device 6 executes the mobile control process shown in FIG. 7 for each control step (each control time), it is shown in FIG. 8 at a predetermined control time (T). as such, a plurality of travel patterns Cs (vs, ωs) (s = 1,2, ..., M S) to select a specific movement pattern from among outputs a movement command C (v, ω).

The plurality of movement patterns Cs are prepared in advance in combination with an arbitrary translational velocity vs. an arbitrary angular velocity ωs. In the example of the movement pattern Cs shown in FIG. 8, five movement patterns C _{1 to} C ₅ are prepared, the translational velocity vs. is fixed, and the angular velocity ωs is changed in five steps. The plurality of movement patterns Cs may be combined with those in which the translational velocity vs. are changed, or those in which both the translational velocity vs. the angular velocity ωs are changed may be combined.

Further, when calculating the evaluation value J, the moving body control device 6 has a plurality of movement patterns Cs (T + t) with respect to the control time (T + M _{t ) ahead of the control time (T) to the predetermined number of control steps M t.} A specific movement pattern is selected using _{a plurality of time-series movement patterns D k} = {Cs (T + 1), (T + 2), ..., Cs (T + M _t )}, which are a combination of the above in a tree structure in time series. In the example of time-series moving pattern D ₁ shown in FIG. 8, between the control time (T + 1) to control time (T + M _t), it is a combination of movement patterns C ₁ rectilinear, the time-series moving pattern D ₁ = It is configured as {C ₁ (T + 1), C ₁ (T + 2), ..., C ₁ (T + M _t )}. In the plurality of time-series movement patterns D _k (k = 1, 2, ..., M _k ), for example, the number of patterns M _s of the plurality of movement patterns Cs is 5, and the number of control steps M _t is 4. If it is round, the number of combinations M _k of the plurality of time-series moving pattern D _k would "5 ^4" number is present.

In a mobile control processing shown in FIG. 7, first, the moving body information obtaining unit 610 obtains mobile body information 600 including the position _{P r} and the velocity vector _{v r} of the target position _P ^{r goal} and the robot R of the robot R ( Step S10). Here, the mobile information acquisition means 610 reads the mobile information 600 from the storage unit 60, but the mobile information 600 may be acquired directly from the own position detection unit 3.

Then, the pedestrian information acquiring unit 611 acquires pedestrian information 601 including the position _{P hi} and velocity vector _{v hi} pedestrian _{H i} present in a detection range 40 of the robot R (step S20). Here, the pedestrian information acquisition means 611 reads the pedestrian information 601 from the storage unit 60, but the pedestrian information 601 may be directly acquired from the pedestrian detection unit 4 or via the communication unit 62. Pedestrian information 601 may be acquired from an external device.

Next, the evaluation value calculation means 612 initializes the variable k to “1” (step S30), and selects the time-series movement pattern D _{k to be processed from the plurality of time-series movement patterns D k} (step S31). ).

Next, the evaluation value calculation means 612 initializes the variable t to “1” (step S32) and selects the movement pattern Cs (T + t) of the processing target included in _{the time-series movement pattern Dk of the processing target (step).} S33). Then, it is assumed that the robot R moves according to the movement pattern Cs (T + t) and the position and velocity vector of the robot R is changed. pedestrian information (position and velocity vector of the pedestrian _{H i)} predict the (updated) (step S34). Further, the evaluation value calculation unit 612, based on the predicted moving body information and pedestrian information, calculates an evaluation value J _t by the above equation (1) (step S35).

Next, the evaluation value calculating means 612 increments the variable t (step S36: Yes), returns to step S33, and then changes the movement pattern Cs (T + t) to be processed, and then steps S34 and S35 described above. _{Is repeated M t} times for the number of control steps, and then the loop processing of the variable t is exited (step S36: No).

_{Then, the evaluation value calculation means 612 adds a plurality of evaluation values J t} (t = 1, 2, ... M _t ) at each control time calculated in chronological order in step S35, that is, the above. (2) below, to calculate the overall evaluation value Js _k for the time-series moving pattern _{D k} to be processed (step S37). It should be noted that, omitting the calculation of the evaluation value _{J t} at step 35, may be only for calculating the overall evaluation value Js _k at step S37.

Here, as described above, by the steps S33~S35 are repeated by the loop processing of the variable t, the state of the robot R and pedestrians H _i at each control time (mobile information and pedestrian information)) Cumulative _{The evaluation value Jt} is calculated based on the result predicted (updated) in. Therefore, the overall evaluation value Js _k calculated in step S37, the environment when the robot R moves in time series moving pattern D _k, i.e., the previous from the current control time (T) to control the number of steps M _t The effect of the robot R on the pedestrian H approaching the robot R side (first virtual repulsive force) at each control time of the control time (T + M _{t), and the pedestrian H exerting on other pedestrians H.} The effect (second virtual repulsion) is cumulatively reflected.

Next, the evaluation value calculation means 612 increments the variable k (step S38: Yes), _{then changes the time-series movement pattern D k} to be processed, and performs the above steps S31 to S37 in the time-series movement pattern. After repeating the combination number of D _{k for M k} , the loop processing of the variable k is exited (step S38: No).

Next, the movement control unit 613, a plurality of total evaluation values calculated at steps _{S37 Js k (k = 1,2,} ..., M k) by comparing, for a plurality of time-series moving pattern _{D k} A specific time-series movement pattern D corresponding to the minimum comprehensive evaluation value is selected from among them (step S40). Then, the movement control means 613 moves by outputting the movement command C (v, ω) according to the movement pattern Cs (T + 1) of the earliest control step included in the time-series movement pattern D to the movement unit 5. The unit 5 is controlled (step S41).

As described above, the mobile body control device 6 executes the mobile body control process shown in FIG. 7 at predetermined control cycles, and controls the mobile unit 5 based on the movement command C (v, ω). , move the robot R to the target position _P ^{r goal} while avoiding a collision with a pedestrian _{H i.}

(Analysis result by simulation)
Next, the result of analyzing the situation in which the robot R controlled by the moving body control device 6 moves by simulation will be described.

(Comparison with and without the second virtual attraction)
FIG. 9 is a diagram showing an analysis result simulating the movement state of the robot R when the second virtual attractive force is set to “none”. FIG. 10 is a diagram showing an analysis result simulating the movement state of the robot R when the second virtual attractive force is set to “Yes”.

As shown in FIG. 9, when the second virtual attractive force is set to "none", the robot R is a four pedestrians H _{1 to} H ₄ (group 1) coming from the front of the robot R toward the robot R side. ) Pass by on the left side.

On the other hand, as shown in FIG. 10, when the second virtual attractive force is set to "Yes", two people who are in front of the robot R and head in the same direction as the robot R when calculating the evaluation value J A second virtual attraction that acts to attract the robot R to the pedestrians H ₅ and H _{6 (group 2) is considered.} Therefore, the robot R, as shown in FIG. 10 (a), (b), so as to follow the two pedestrian _H 5, _{H 6} (Group 2), and diverted to the right front, the robot R pass each other pedestrian H ₁ to H 4 from the front of the four coming towards the robot R side so as to pass through the right side of _(group 1).

FIG. 11 shows the movement status of the robot R between the case where the second virtual attraction is set to “none” (FIG. 9) and the case where the second virtual attraction is set to “yes” (FIG. 10). It is a figure which shows the comparison result of comparison.

If the second virtual attraction is "no", in order to directly facing the robot R transgressions 4 pedestrians _H 1 _{~H 4} (group 1), 4 pedestrians _H 1 _{~H 4} (group 1 ) Receives a large repulsive force (influence) from the robot R. On the other hand, when the second virtual attraction is "Yes", when viewed from four of the pedestrian _H 1 _{~H 4} (group 1), the robot R, 2 pedestrians _H 5, _{H 6} (Group to position so as to be hidden behind the 2), 4 pedestrians H ₁ to H 4 _(group 1) repulsive force (impact) is reduced to receive from the robot R. As a result, the robot R _{follows two pedestrians H 5} and H ₆ (group 2), _{thereby reducing the influence on other pedestrians H 1 to} H ₄ (group 1) and walking around. It turned out that it can move in harmony with the person.

(Comparison with and without the second virtual repulsion)
FIG. 12 is a diagram showing an analysis result simulating the movement state of the robot R when the second virtual repulsive force is set to “none”. FIG. 13 is a diagram showing an analysis result simulating the movement state of the robot R when the second virtual repulsive force is set to “Yes”.

As shown in FIG. 12, when the second virtual repulsive force is set to "None", the robot R is a four pedestrians H _{1 to} H ₄ (4 pedestrians H 1 to H 4) approaching the robot R side from a slightly left front of the robot R. group 1), pass each other to divide between the robot 1 pedestrians H slightly coming toward the right front to the robot R side of the R _{5 (group} 2).

On the other hand, as shown in FIG. 13, when the second virtual repulsive force is set to "Yes", each of the _{five pedestrians H 1 to} H 5 heading toward the robot R side and others when calculating the evaluation value J, etc. pedestrian (e.g., other pedestrian against pedestrians H _1, pedestrian H ₂ to H ₅ of four applicable.) second virtual repulsive force acting between the are considered. Therefore, the robot R acts between pedestrians by dividing between _{four pedestrians H 1 to} H 4 (group 1) and one pedestrian H _{5 (group 2).} as the repulsive force is not increased, to avoid the pedestrian H ₁ to H ₅ of 5 people, and diverted to the right front, pass each other so as to pass through the five right pedestrian H ₁ to H _5.

FIG. 14 shows the movement status of the robot R between the case where the second virtual repulsive force is set to “none” (FIG. 12) and the case where the second virtual repulsive force is set to “yes” (FIG. 13). It is a figure which shows the comparison result of comparison.

If the second virtual repulsive force is "None", passing the robot R, and 4 pedestrians H ₁ to H 4 _(Group 1), between 1 pedestrians H _{5 (Group} 2) in time period before and after (FIG. 12 (b)), pedestrian _H 1 to H ₅ of five repulsion (impact) force is increased to receive from the robot R.

On the other hand, if the second virtual repulsive force is "present", when viewed from 5 pedestrians H ₁ to H _5, for changing the path to the robot R to avoid spontaneous, five walking The force that the persons H _{1 to} H ₅ receive from the robot R becomes smaller. Thus, the robot R is to reduce the impact on the pedestrian H ₁ to H _5, was found to be moving in harmony with the surrounding pedestrian.

(Other embodiments)
Although the embodiments of the present invention have been described above, the present invention is not limited to the above embodiments and can be appropriately modified without departing from the technical idea of the present invention.

For example, in the above embodiment, the moving body 1 has been described as being a self-propelled robot R, but the moving body 1 is not limited to the self-propelled robot R, for example, a self-propelled transport trolley. Or a vacuum cleaner.

Further, in the above embodiment, the moving unit 5 of the moving body 1 has been described as having a plurality of wheels 50, but the moving unit 5 may have a plurality of wheels 50, for example, an omni wheel. Instead of the plurality of wheels 50, a walking mechanism for bipedal walking may be provided.

Further, in the above embodiment, the mobile control device 6 has been described as using the time-series movement pattern shown in FIG. 8, but a single movement pattern may be used, and in that case, it is shown in FIG. In the flowchart (moving body control process), the _{process may be performed with M t} = 1 and the evaluation value J may be calculated.

Further, in the above embodiment, the moving body 1 has been described as including the moving body control device 6, but the moving body control device 6 is configured to be able to communicate various information with the moving body 1 so that the moving body 1 can communicate with the moving body 1. It may be provided in a device separate from 1.

Further, in the above embodiment, the mobile control program 603 has been described as being stored in the storage unit 60, but it can be read by a computer such as a CD-ROM or DVD in an installable format or an executable format file. It may be recorded and provided on various recording media. Further, the mobile control program 603 may be provided by storing it on a computer connected to a network such as the Internet and downloading it via the network.

1 ... mobile body (robot R), 2 ... main body, 3 ... self-position detection unit,
4 ... Pedestrian detection unit, 5 ... Moving unit (means of transportation),
6 ... mobile control device, 7 ... power supply,
40 ... detection range, 50 ... wheels, 51 ... electric motor,
60 ... storage unit, 61 ... control unit, 62 ... communication unit,
600 ... mobile information, 601 ... pedestrian information,
602 ... Setting information, 603 ... Mobile control program,
610 ... Mobile information acquisition means, 611 ... Pedestrian information acquisition means,
612 ... Evaluation value calculation means, 613 ... Movement control means,
_{H, H i, H 1 ~H} 6 ... pedestrian

Claims (7)

A moving body control device that controls the moving means of the moving body so as to move the moving body to a predetermined target position in an environment where a pedestrian exists.
A moving body information acquisition means for acquiring moving body information including the target position, the position of the moving body, and a velocity vector, and
A pedestrian information acquisition means for acquiring pedestrian information including the position and speed vector of the pedestrian existing around the moving body, and
An evaluation value calculating means for calculating a plurality of evaluation values for each of the plurality of movement patterns for the moving means based on the moving body information and the pedestrian information.
A movement control means for selecting a specific movement pattern from the plurality of movement patterns by comparing the plurality of evaluation values and controlling the movement means based on a movement command according to the movement pattern is provided.
The evaluation value calculation means is
A first virtual attraction that acts to attract the moving object to the target position,
A second virtual attraction that exists in front of the moving body and acts to attract the moving body to the pedestrian heading in the same direction as the moving body.
A first virtual repulsive force acting between the pedestrian and the moving body toward the moving body side,
The evaluation value is calculated by synthesizing a second virtual repulsive force acting between the pedestrian toward the moving body side and another pedestrian.
A mobile control device characterized by the fact that. The evaluation value calculation means is
When synthesizing the second virtual attraction
The more the pedestrian who exists in front of the moving body and heads in the same direction as the moving body walks toward the target position, the more the pedestrian acts between the pedestrian and the moving body. The second virtual attraction is weighted so that the second virtual attraction becomes large.
The mobile control device according to claim 1. The evaluation value calculation means is
When synthesizing the second virtual attraction
Weighting is performed on the second virtual attraction in consideration of the relative speed between the pedestrian existing in front of the moving body and heading in the same direction as the moving body and the moving body. ,
The mobile control device according to claim 1 or 2, wherein the mobile control device is characterized by the above. The evaluation value calculation means is
When synthesizing the first virtual repulsive force
The first virtual repulsive force acting between the pedestrian and the moving body increases as the pedestrian heading toward the moving body walks toward the moving body. Weighting the virtual repulsive force of
The mobile control device according to any one of claims 1 to 3, wherein the mobile control device is characterized by the above. The evaluation value calculation means is
For each of the plurality of time-series movement patterns in which a plurality of movement patterns are combined in time series with respect to the control time ahead from the predetermined control time to the predetermined number of control steps, the plurality of movement patterns included in the time-series movement pattern. The moving body information and the pedestrian information when it is assumed that the moving body has moved according to the movement pattern are cumulatively predicted, and the evaluation at each control time based on the predicted moving body information and the pedestrian information. By adding the values, the comprehensive evaluation value is calculated respectively.
The movement control means
A specific time-series movement pattern is selected from the plurality of time-series movement patterns by comparing the plurality of comprehensive evaluation values, and the movement pattern of the earliest control step included in the time-series movement pattern is used. Control the moving means based on the moving command,
The mobile control device according to any one of claims 1 to 4, wherein the mobile control device is characterized by the above. The computer functions as the mobile control device according to any one of claims 1 to 5.
A mobile control program characterized by this. The mobile control device according to any one of claims 1 to 5 is provided.
A mobile body characterized by that.

Images