JP5160311B2 - Autonomous mobile device and control method of autonomous mobile device - Google Patents

Description

  The present invention relates to an autonomous mobile device that is equipped with a control device and performs autonomous movement under the control of the control device, and a control method for such an autonomous mobile device.

  2. Description of the Related Art Conventionally, an autonomous mobile device that is equipped with a control device and performs autonomous movement under the control of the control device has been proposed. The autonomous mobile device moves on a designated movement route in advance corresponding to map information, or moves on a designated movement route based on a remote operation signal designating the movement route.

  In order to make the autonomous mobile device reach the destination, the control device in such an autonomous mobile device measures the presence or absence of an obstacle in the outside world by using a sensor, determines whether or not the vehicle can move, and corrects the movement route. In other words, when there is an obstacle on the movement route while following the designated movement route, the control device once deviates from the designated movement route to avoid the obstacle, and removes the obstacle. After avoiding, control is performed to return to the designated travel route.

  In such an autonomous mobile device, in order to measure the presence or absence of an obstacle on the moving route, using a sensor such as a laser range finder (LRF) or a stereo camera, obtaining shape information from information obtained from these sensors, Judging whether there is an obstacle. Thereby, autonomous movement while avoiding an obstacle becomes possible.

  By the way, in a passenger-driven device such as a passenger car, when passing through an intersection with poor visibility, the driver first recognizes that there is a portion with poor visibility, and then stops and looks at it. After confirming that there are no moving obstacles such as other devices, the movement is resumed. Also, on such roads, there are generally signs such as traffic lights and temporary stops, and the driver ensures safety by driving according to these signs.

  In an autonomous mobile device, if there is a moving obstacle within the measurement range of the sensor, it automatically tracks the moving obstacle, prefetches the movement path of the moving obstacle, and avoids a collision in advance. Yes. However, since autonomous mobile devices cannot recognize that there are places with poor visibility in places with poor visibility, such as intersections, it is not possible to take action to secure the visibility, such as temporarily stopping, and moving obstacles The presence or absence of an object cannot be fully confirmed. Further, the autonomous mobile device cannot recognize a signal or a sign such as a temporary stop.

  As a technique for realizing a collision avoidance operation at such an intersection or the like, for example, in Patent Document 1, based on GPS information and map information, the operation of a blinker, the presence or absence of a pedestrian crossing, and the like are detected. A technique for automatically estimating the existence of intersections and moving obstacles is described.

  Patent Document 2 discloses a technique for securing safety in a region with poor visibility by attaching a transmitter in advance to a moving obstacle such as a pedestrian and detecting a signal emitted by the transmitter in an autonomous mobile device. Is described.

Japanese Patent No. 3617307 Japanese Patent Laid-Open No. 11-232669

  The conventional moving obstacle avoidance method in an autonomous mobile device requires that all moving obstacles exist within the measurement range of the sensor. Therefore, the conventional autonomous mobile device cannot cope with a moving obstacle that pops out from the blind spot.

  And when applying the conventional technique for implement | achieving the collision avoidance operation | movement in an intersection etc. to an autonomous mobile device, the following problems can be considered.

  That is, in order to be able to recognize signals, signs, etc., in order to automatically detect these signals, etc., a dedicated detection device for performing image processing, etc. is required, and the device configuration becomes complicated and large. End up.

  Even if a signal or sign can be recognized, the danger of a collision cannot be avoided on a road or environment where no signal or sign is installed.

  In addition, as described in Patent Document 1, in the technology using GPS information and map information, moving obstacles that exist in places with poor visibility that are not in known information, such as the shadow of a vehicle parked on the road. Can not avoid the danger of collision. Even if an intersection is detected by detecting a pedestrian crossing or the like, the risk of a collision cannot be avoided on a road without a pedestrian crossing or the like.

  Furthermore, as described in Patent Document 2, in the technique of attaching a transmitter to a moving obstacle, it is necessary to attach a transmitter to all the moving obstacles, which cannot be executed in any environment.

  Therefore, the present invention is proposed in view of the above-described circumstances, and is an autonomous mobile device and a method for controlling the autonomous mobile device, where an unmeasured region (dead angle) due to a wall or the like exists and is not carelessly unknown. When moving in an area where a moving obstacle can pop out, the collision of the encounter is autonomously avoided regardless of the presence of signals, signs, pedestrian crossings, etc., without complicating and increasing the size of the device configuration. It is an object of the present invention to provide an autonomous mobile device and a method for controlling the autonomous mobile device.

  In order to solve the above-described problems and achieve the object, the present invention has any one of the following configurations.

[Configuration 1]
An autonomous mobile device according to the present invention includes an unmanned mobile device that can be moved, a sensor that is mounted on the unmanned mobile device and measures the presence of fixed obstacles and moving obstacles around the unmanned mobile device, and the unmanned mobile device Based on the measurement result of the sensor mounted on the vehicle, a movable area without a fixed obstacle is detected, a planned movement path of the unmanned mobile device is determined in this movable area, and the unmanned mobile device is controlled to move on the planned movement path Control means for detecting a moving obstacle based on a measurement result of the sensor and predicting a movement path of the moving obstacle to avoid a collision with the moving obstacle. orthogonally from any point above distance of the fixed obstacle, or, if the boundary and the movable region and the non-measurement area by the moving obstacle contact is found, from any point to the boundary There, except where it is a certain distance enough to accommodate a collision avoidance operation after moving obstacle from unmeasured region flew out, and with the possibility segment pop-out the start position of the search, possibly interval popping viewed from the unmanned mobile device Brake is performed so that the vehicle can be decelerated or stopped by the start position, and when a moving obstacle is detected, the stop state is continued, and when there is no moving obstacle, the speed returns to the original speed, The movement along the planned movement path is resumed.

[Configuration 2]
The method for controlling an autonomous mobile device according to the present invention controls the movement of the unmanned mobile device based on a measurement result by a sensor mounted on a movable unmanned mobile device and measuring the presence or absence of a surrounding fixed obstacle and a moving obstacle. A method for controlling an autonomous mobile device, in which a movable area without a fixed obstacle is detected based on a measurement result by a sensor, a planned movement path of an unmanned mobile device is determined in the movable area, and the measurement result by the sensor Based on the detection of the moving obstacle and predicting the moving path of the moving obstacle, controlling the unmanned moving device to avoid the collision with the moving obstacle, moving on the planned moving path, and arbitrarily moving on the planned moving path. perpendicularly from the viewpoint of a fixed obstacle, or, if the boundary and the movable region and the non-measurement area by the moving obstacle contact is found, the unmeasured territory distance from any point to the boundary Except when moving obstacle is the distance a margin that may collision avoidance operation after jumped out, and the search of start position from jumping out potential section, in to a position where potential section popping viewed from the unmanned mobile device begins If the vehicle is detected so that it can be decelerated or stopped and a moving obstacle is detected, it will continue to stop, and if there is no moving obstacle, it will return to its original speed and enter the planned movement route. The movement along the line is resumed.

In the autonomous mobile device according to the present invention having the configuration 1, the control means includes a fixed obstacle or an unmeasured area and a movable area due to the moving obstacle in a direction perpendicular to an arbitrary point on the planned movement route. If a touching boundary is found, start the search unless the distance from any point to the boundary is a distance that allows a collision avoidance action after a moving obstacle has popped out of the unmeasured area Set the position as a pop-out possibility section, brake it so that it can decelerate or stop until the position where the pop-out possibility section starts when viewed from the unmanned mobile device, and if a moving obstacle is detected, continue the stop state, When there are no moving obstacles, the speed returns to the original speed and the movement along the planned movement path is resumed, so there is an unmeasured area (blind spot) and the moving obstacles pop out inadvertently. In the case of moving the so that a region, it is possible to autonomously avoid collision Crossing.

  Therefore, this autonomous mobile device can avoid collisions reliably by moving at high speed in areas where there is no possibility of jumping out of moving obstacles and decelerating in areas where there is a possibility of jumping out. It is.

In the control method of the autonomous mobile device according to the present invention having the configuration 2, the fixed obstacle or the unmeasured area due to the moving obstacle and the movable area are in contact with each other at a right angle from any point on the planned movement route. If a boundary is found , the search start position , except when the distance from any point to the boundary is a distance that allows a collision avoidance operation after a moving obstacle has popped out of the unmeasured area If the obstacle is detected by moving to a position where it can be decelerated or stopped until it reaches the position where the pop-out possibility section starts when viewed from the unmanned mobile device, the stop state is continued and moved. If there are no obstacles, the speed will return to the original speed, and the movement along the planned route will be resumed, so there will be an unmeasured area (dead zone) and the moving obstacle will pop out inadvertently. In the case of moving the area as well, it is possible to autonomously avoid collision Crossing.

  Therefore, according to the control method of the autonomous mobile device, the collision is surely performed by moving at a high speed in an area where a moving obstacle is not likely to jump out and decelerating in an area where the obstacle is likely to jump out. It is possible to avoid it.

  That is, the present invention is an autonomous mobile device and an autonomous mobile device control method, in the case where there is an unmeasured area (dead angle) due to a wall or the like, and when moving an area where an unknown moving obstacle can pop out carelessly, An autonomous mobile device and an autonomous mobile device control method capable of autonomously avoiding a collision at an encounter, regardless of whether a signal, a sign, a pedestrian crossing, or the like is installed without complicating or increasing the size of the device It can be provided.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

[Configuration of autonomous mobile device]
FIG. 1 is a side view showing an external configuration of an autonomous mobile device according to the present invention.

  As shown in FIG. 1, the autonomous mobile device includes an unmanned vehicle (UGV) 1 that can be moved, and a control unit (autonomous control computer) 2 mounted on the unmanned vehicle 1. . The unmanned vehicle 1 is controlled by the control means 2. This control means 2 enables autonomous movement of the unmanned vehicle 1 based on the map information.

  The unmanned vehicle 1 is equipped with a sensor 3 for measuring surrounding fixed obstacles and moving obstacles, a GPS receiver 4 for self-position measurement, an odometer, a vertical gyro for posture measurement, and the like. As the sensor 3, for example, a laser range finder (LRF), a stereo camera serving as a distance measuring sensor, or the like can be used.

  FIG. 2 is a block diagram showing a control hardware configuration in the autonomous mobile device according to the present invention.

  The control means 2 of this autonomous mobile device comprises a sensor data input computer 2a, a movable area determination / map generation computer 2b, and a machine control processing computer 2c. Each computer 2a, 2b, 2c is connected by LAN10 and communicates each processing result.

  Data from the external measuring device 5 and the self-position measuring device 6 are input to the sensor data input computer 2 a of the control means 2 through the I / O bus (data bus) 8. The movable area determination / map generation computer 2b determines whether or not movement of a predetermined movement route is possible based on the data and map information input to the sensor data input computer 2a. The airframe control processing computer 2c is connected to a motor driver and a D / A converter through an I / O bus 9, and controls a vehicle body control device 7 including a handle actuator, a brake, an accelerator, a by-wire, and the like to move the unmanned vehicle 1 Control.

  FIG. 3 is a block diagram showing a control software configuration in the autonomous mobile device according to the present invention.

  In the control software configuration in the control means 2 of the autonomous mobile device, a plurality of sensor data I / O modules, that is, an LRF data I / O module 11, a self-position data I / O module 12, and an airframe control I / O module 13 are used. have. In these sensor data I / O modules, pre-processing such as coordinate conversion is performed for the laser range finder (LRF) and data integration is performed for the self-position. The self position information generation module 14 generates self position information of the unmanned vehicle 1. The obstacle detection module 19 measures the presence or absence of surrounding fixed obstacles and moving obstacles based on the sensor data from the LRF data I / O module 11, generates obstacle / movable area information, and Send to generation module 15.

  Then, the map generation module 15 maps the movement availability information based on the data from the laser range finder (LRF) obtained while moving on a single grid-like map, and displays the obstacle / movable area / A map showing the measurement area is generated.

  The movement route generation module 17 generates the planned movement route 106 based on the created map. The pop-out area detection module 18 detects the boundary between the movable area and the unmeasured area, that is, the “pop-out area”, and detects the distance between the “pop-out area” and the planned movement path 106. In addition, the aircraft control module 16 determines the amount of operation of each actuator to move along the determined planned movement path 106 and instructs the aircraft control I / O module 13 on the steering wheel angle / vehicle speed.

[Operation of autonomous mobile device]
FIG. 4 is a plan view showing a moving state when there is a moving obstacle of the autonomous mobile device according to the present invention.

  As shown in FIG. 4, the sensor 3 provided in the autonomous mobile device according to the present invention is a sensor capable of ranging, such as an LRF or a stereo camera, and is specified by a predetermined measurement range (angle range and distance). The presence or absence of fixed obstacles and moving obstacles in the area).

  This autonomous mobile device detects the position and moving direction of the moving obstacle 101 when the moving obstacle 101 enters the sensor measurement range, and avoids a collision with the moving obstacle 101 based on the detection result. The turning / braking control is performed.

  FIG. 5 is a plan view showing a situation with poor visibility, such as an intersection on a road surrounded by walls.

  As shown in FIG. 5, in a situation with poor visibility such as an intersection on a road surrounded by a wall 102, the robot moves to a position that would normally be within the sensor measurement range (within a predetermined angle range and within a predetermined distance). Even if the obstacle 101 exists, the presence of an obstructing object such as a fixed obstacle (such as a wall) 102 causes a blind spot in the sensor 3 and makes it impossible to detect the moving obstacle 101.

  FIG. 6 is a plan view showing a measurement state by the sensor in a situation where visibility is poor.

  As shown in FIG. 6, the wall 102 and the like can be detected as fixed obstacles, and the autonomous mobile device moves in the movable region while avoiding these fixed obstacles. Since the inside of the intersection is detected as the movable region 103, the conventional autonomous mobile device enters the intersection without turning or braking. At this time, there is a possibility of colliding with the moving obstacle 101 in the intersection.

  FIG. 7 is a plan view showing an example of a map used for autonomous control movement control in a situation with poor visibility.

  That is, as shown in FIG. 7, the conventional autonomous mobile device sequentially determines sensor measurement results, distinguishes between the movable area 103 and the fixed obstacle 102, and a planned movement path that continuously connects the movable areas. 106 is automatically generated and moved on this route.

  FIG. 8 is a plan view showing the principle of detection of a region with a possibility of popping out.

  On the map created from the sensor measurement result, there is a boundary 105 where the movable area 103 and the unmeasured area 104 are in contact with each other, as shown in FIG. This occurs due to the presence of a fixed obstacle 102 that blocks the line of sight. At these boundaries 105, there is a possibility that the movable area 103 extends in the direction of the unmeasured area 104. Since the movable area 103 is an area that can be moved by the moving obstacle 101, there is a possibility that the moving obstacle 101 that moves in the unmeasured area 104 exists.

  However, the moving obstacle 101 existing in the movable area 103 blocked by the fixed obstacle 102 cannot be measured from the sensor 3 of the autonomous mobile device.

  Therefore, by detecting the boundary 105 where the movable area 103 and the unmeasured area 104 are in contact with each other, it is possible to automatically detect an area that may be popped out.

  (A) in FIG. 9 is a plan view showing a method for detecting a region that may jump out, and (b) in FIG. 9 is a plan view showing a map that is actually detected.

  In this autonomous mobile device, as shown in (a) of FIG. 9, the attribute on the map is perpendicular to the arbitrary point (p0, p1,..., Pi) on the planned movement path 106 on the map. To evaluate. At this time, if a location (boundary 105) that changes from the movable region 103 to the unmeasured region 104 is found, the search start position (p0, p1,..., Pi) is stored.

  At this time, the distance from the search start position to the boundary 105 may be sufficiently large, such as when the road is wide. In this case, in consideration of the distance and the general moving speed of the moving obstacle 101, it is determined that the distance that allows a normal collision avoidance operation after the moving obstacle 101 jumps out of the unmeasured area 104 is determined. If so, the collision avoidance process described below is not performed.

  After such a search is completed, the stored search start position on the planned movement route 106 becomes a pop-out possibility section 106a that may pop out.

  When the horizontal measurement angle range of the sensor 3 is smaller than 180 °, the movable area 103 and the unmeasured area 104 are not included in the area near the autonomous mobile device even if the fixed obstacle 102 is not present. A boundary 105 that contacts is generated. Since such a boundary 105 is ignored, the boundary 105 is not detected in the vicinity region where the distance from the autonomous mobile device is equal to or less than a predetermined distance. The range of such a nearby region is determined based on the measurement angle range and road width of the sensor 3, the assumed speed of the moving obstacle 101, and the like.

  Further, even when the planned movement route 106 is curved, as described above, a search in the direction perpendicular to the planned movement route 106 is possible. However, since the search density is sparse outside the arc, in this case, it is preferable to perform processing such as widening the search width as it becomes farther away.

  When an actual sensor is used, a dense map as shown in FIG. 9A may not be obtained. In this case, it is possible to remove the island-like unmeasured area and noise by the median filter and the morphological process while holding the obstacle area. As shown in FIG. A map with no problems is obtained. In FIG. 9B, the black portion indicates the obstacle region (road shoulder), and the white portion indicates the movable region 103.

  FIG. 10 is a plan view showing the control of the autonomous mobile device according to the present invention after the pop-out occurrence section is detected.

  As shown in FIG. 10, in the autonomous mobile device, when a pop-out possibility section 106a having a possibility of popping out occurs on the planned movement route 106, the autonomous mobile device reaches the position where the pop-out possibility section 106a starts when viewed from the vehicle body. , Brake so that you can slow down or stop.

  FIG. 11 is a plan view showing a state when the autonomous mobile device according to the present invention approaches the popping possibility section.

  As shown in FIG. 11, the autonomous mobile device according to the present invention decelerates or stops when approaching the pop-out possibility section 106 a, so that it does not collide with another moving obstacle 101 by itself. As it approaches the pop-out possibility section 106a, the blind spot is reduced, and a wider area can be measured.

  In addition, when the measurement angle range of the sensor 3 is narrower than 180 °, the measurement may be performed all over by pan control of the sensor 3 or the like. At this time, since the autonomous mobile device is decelerated or stopped, there is no problem even if measurement takes time.

When the moving obstacle 101 is detected, the stop state is continued. When the moving obstacle 101 does not exist, and returns to the original speed, resumes movement along the movement scheduled route 106.

  FIG. 12 is a plan view showing a state in which the autonomous mobile device according to the present invention passes through a blind corner (such as a curved saddle road).

  When the autonomous mobile device according to the present invention passes through a blind corner (curved road, etc.), as shown in FIG. 12, the pop-out possibility section 106a is deeper in the traveling direction than the intersection or the like. Become. Further, in the curve, since the blind spot is sequentially decreased as the autonomous mobile device progresses, the collision can be avoided by starting the braking control with the jumping possibility section 106a as a target.

  FIG. 13 is a flowchart showing an operation of searching for a pop-out possibility section.

  As shown in FIG. 13, the control means 2 of this autonomous mobile device starts operation in step st1, and when it proceeds to step st2, as shown in (a) in FIG. Search start positions (p0, p1,..., Pi) are defined. At this time, p0 is the nearest position in front of the autonomous mobile device. This is based on data measured and held in the past when the autonomous mobile device approaches the pop-out possibility section 106a and decelerates or stops, and the measurement range of the sensor 3 is 180 °. This is to use a map. That is, if the data measured in the past is held, even if the measurement range of the sensor 3 is 180 ° or less, as shown in FIG. Because there exists. The diagram shown on the right side of (b) in FIG. 9 is an obstacle map created based on the actually acquired data, because the autonomous mobile device moves to a point where data was acquired from a distance in the past. Even if the measurement range of the sensor 3 is not 180 °, there is a measurement result on the map for the nearest position of the autonomous mobile device.

  Next, the process proceeds to step st3, and a boundary 105 between the movable area 103 and the unmeasured area 104 is searched independently of the planned movement path 106 in the left-right direction. Since the search is performed independently in the left-right direction, it is possible to deal with a case where there is a fixed obstacle 102 such as a wall on only one of the left and right sides.

  It progresses to step st4, it is discriminate | determined whether the boundary 105 was detected, and if the boundary 105 is detected, it will progress to step st5, and if the boundary 105 is not detected, it will progress to step st6.

  In step st5, the length L from the search start position to the boundary 105 between the movable area 103 and the unmeasured area 104 and the left or right side of the boundary 105 are recorded, and the process proceeds to step st6. This length L is used to determine whether or not the path on which the autonomous mobile device moves is sufficiently wide in the processing described later.

  In step st6, it is determined whether pi has reached the end point of the search. If it is not the end point, the process proceeds to step st7, and if it is the end point, the process proceeds to step st8. The search end point is a limit point (farthest point) of the measurement range of the sensor 3 on the planned movement path 106.

  In step st7, pi is advanced by one (added), and the process returns to step st3. That is, the search is performed while the search start position is sequentially advanced. When the width of the assumed moving obstacle 101 is A, the interval to be advanced is less than half of that. For example, the width of the moving obstacle 101 is about 0.5 m for a person and about 1.5 m for a vehicle.

  Steps after step st8 are processing for dealing with cases where the pop-out possibility section 106a is narrow and wide. In step st8, in order to start checking the continuity of the boundary 105, the process proceeds to step st9.

  In step st9, it is determined whether the boundary 105 is searched to the side of the planned movement route 106. If the boundary 105 is searched, the process proceeds to step st10, and if the boundary 105 is not searched, the process proceeds to step st13.

  In step st10, the continuity of the boundary 105 is examined in the vicinity of the place where the boundary 105 is detected, and the process proceeds to step st11.

  In step st11, the continuity of the boundary 105 between the movable area 103 and the unmeasured area 104 recorded on the planned movement path 106 at the search start position pi is greater than the width A of the moving obstacle 101 ( It is determined whether or not it has a road width. If the continuity of the boundary 105 is equal to or greater than the width A of the moving obstacle 101, the process proceeds to step st12. If the boundary 105 is not equal to or larger than the width A of the moving obstacle 101, it is determined that the moving obstacle 101 cannot come out. Therefore, it is determined that it is not the pop-out possibility section 106a, and the process proceeds to step st13.

  In step st12, the pop-out possibility section 106a is recorded, and the process proceeds to step st13.

  In step st13, it is determined whether pi is the end point of the continuity check of the boundary 105. If it is not the end point, the process proceeds to step st14, and if it is the end point, the process proceeds to step st15 to end the process.

  In step st14, pi is advanced by one (added), and the process returns to step st9.

  FIG. 14 is a flowchart illustrating the operation of the pop-out possibility section detection module.

  The process shown in FIG. 14 is the details of the process in the pop-out area detection module 18 shown in FIG. This control method is a process aimed at avoiding a collision with the moving obstacle 101 existing in front of the autonomous mobile device. A situation in which a rear-end collision is performed at a higher speed than the autonomous mobile device from behind the autonomous mobile device is not considered.

  As shown in FIG. 14, when the process is started in step st16, the process proceeds to step st17, the map and the data of the planned movement route 106 are acquired, and the process proceeds to step st18. In step st18, as described above, the pop-out possibility section 106a is searched, and the process proceeds to step st19.

  In step st19, it is determined whether or not a pop-out possibility section 106a exists on the scheduled movement route 106. If there is a pop-out possibility section 106a, the process proceeds to step st20, and if there is no pop-out possibility section 106a, a step is performed. Proceed to st22.

  In step st20, the target speed Vd of the autonomous mobile device is determined based on the length L from the planned movement path 106 recorded in step st5 of FIG. 13 to the boundary 105 between the movable area 103 and the unmeasured area 104. Then, the process proceeds to step st21. The target speed Vd is a target value for performing control so that the vehicle decelerates to reach a predetermined point and reaches that speed.

  The target speed Vd based on the distance L is determined as follows. In other words, when the distance L is sufficiently large, the planned movement route 106 is sufficiently separated from the boundary 105 where the pop-out may occur, and even when the pop-out occurs, the moving obstacle 101 is moved along the planned movement route 106. It takes a long time to cross. In this case, since the collision can be avoided by turning after the moving obstacle 101 is detected, the speed Vd is not changed from the initial target speed. These can be derived using the control response speed and turning performance of the autonomous mobile device and the speed of the moving obstacle 101 as parameters. For example, assuming a device that moves at 50 km / h (14 m / sec) as the moving obstacle 101, if the distance L is about 30 m, the time required to cross the planned movement route 106 is about 2 even if a jump occurs. Since there are seconds, an avoidance operation by turning is possible during that time.

  When the distance L is relatively short, the target speed Vd is decelerated or slowed down. In other words, the target speed Vd is set to a speed at which the autonomous mobile device can stop before the moving obstacle 101 suddenly jumps out and crosses the planned movement path 106. These can be derived using the control response time and braking performance of the autonomous mobile device and the speed of the moving obstacle 101 as parameters. For example, assuming a vehicle moving at 50 km / h (14 m / sec) as the moving obstacle 101, if the distance L is about 10 m, it is possible to cross the planned movement route 106 within one second when a jump occurs. Therefore, the speed is reduced to a speed at which the autonomous mobile device can stop by that time, for example, about 5 km / h.

  When the distance L is extremely short, the target speed Vd is set to 0 and temporarily stopped. This is a case where the control response time of the autonomous mobile device is finite, so that there is a possibility of collision even if braking is performed after the moving obstacle 101 is detected. For example, assuming that the moving obstacle 101 is a vehicle moving at 50 km / h (14 m / sec), when the distance L is about 2 m, it intersects with the planned movement route 106 in about 0.15 seconds. In such a time, after the mobile obstacle 101 is discovered, the autonomous mobile device cannot be stopped due to a delay in braking control or the like, and thus needs to be stopped in advance.

  These target speeds Vd are independently determined for each of the left and right directions based on the distance L recorded in step st5. In order to ensure safety, a value with the smaller target speed Vd is adopted.

  FIG. 15 is a graph illustrating an example of a function for determining the target speed.

  As shown in FIG. 15, the target speed Vd is set to 0 when the distance L is less than or equal to a predetermined lower limit value, and is set to no deceleration when the distance L is greater than or equal to a predetermined upper limit value. Is set as a function (for example, a linear function) that monotonously increases with respect to the distance L. Since the relationship between the target speed Vd and the distance L depends on the moving speed, braking performance, turning performance, control response time, etc. of the autonomous mobile device, it is not always the relationship expressed by such a linear function. Absent.

  In step st21, a braking speed plan is determined based on the target speed Vd determined in step st20, and the process proceeds to step st22. This is a process for deciding whether to perform gentle braking when it is necessary to decelerate far, and to perform sudden braking when it is necessary to stop in the vicinity.

  Steps st22 to st24 are processes assuming that the measurement range of the sensor 3 is 180 ° or less. That is, in step st22, a speed threshold Vl of about 3 km / h is set in advance, and it is determined whether the current moving speed of the autonomous mobile device is equal to or lower than the speed threshold Vl. If the moving speed is equal to or less than the speed threshold Vl, the process proceeds to step st24, and if the moving speed is not equal to or less than the speed threshold Vl, the process proceeds to step st23.

  In step st23, the pan operation of the sensor 3 is not performed, and the process proceeds to step st25 while maintaining the normal operation.

  In step st24, since the moving speed of the autonomous mobile device is equal to or less than the speed threshold value V1, panning of the sensor 3 is performed, detection is performed while ensuring a measurement range of 180 ° or more, and the process proceeds to step st25. If the movement speed of the autonomous mobile device is less than or equal to the speed threshold value Vl, this sequence is always performed, and the measurement range of the sensor 3 is set to 180 ° or more. In all cases, such as when stopped, the sensor 3 performs detection over a measurement range of 180 ° or more. Therefore, as shown in step st2 of FIG. 13, even if the search start position p0 is set to the nearest position in front of the autonomous mobile device, the pop-out possibility section 106a can be detected without any problem. Even at the restart, the pop-out possibility section 106a can be detected.

  In step st25, the information on the planned movement route 106 and the speed plan is sent to the machine control I / O module 13, and the process returns to step st17.

  In the autonomous mobile device according to the present invention, if the image processing recognition algorithm is sufficiently advanced, the pop-out possibility section 106a is detected based on the image acquired by the camera in the same manner as described above. Is possible.

  In the above-described embodiment, the autonomous mobile device is a vehicle-type unmanned mobile mechanism, but the present invention is not limited to this. That is, the autonomous mobile device in the present invention may be, for example, a walking type unmanned moving mechanism.

  In addition, the method of searching for the pop-out possibility section described above can be configured not only as an autonomous mobile device but also as a device that issues a warning to the driver by being mounted on a manned vehicle operated by the driver. It is.

It is a side view which shows the external appearance structure of the autonomous mobile device which concerns on this invention. It is a block diagram which shows the control hardware constitutions in the autonomous mobile device which concerns on this invention. It is a block diagram which shows the control software structure in the autonomous mobile apparatus which concerns on this invention. It is a top view which shows a movement state when the movement obstruction of the autonomous mobile device which concerns on this invention exists. It is a top view which shows the situation where a line of sight is bad like an intersection in the road surrounded by walls. It is a top view which shows the measurement state by the sensor in the condition where a line of sight is bad. It is a top view which shows the example of the map used for the movement control of the autonomous control in a situation with a bad view. It is a top view which shows the detection principle of the area | region which may jump out. (A) is a top view which shows the detection method of the area | region which may jump out, (b) is a top view which shows the map actually detected. It is a top view which shows control of the autonomous mobile apparatus which concerns on this invention after detecting a jump generation | occurrence | production area. It is a top view which shows a mode when the autonomous mobile apparatus which concerns on this invention approaches the popping possibility area. It is a top view which shows the state in which the autonomous mobile device which concerns on this invention passes a blind corner (curved tunnel etc.). It is a flowchart which shows the operation | movement which searches for a popping possibility area. It is a flowchart which shows operation | movement of the detection module of a popping possibility area. It is a graph which shows an example of the function of target speed determination.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Unmanned vehicle 2 Control means 3 Sensor 4 GPS receiver 101 Moving obstruction 102 Fixed obstruction 103 Movable area 104 Unmeasured area 105 Boundary 106 Expected movement path 106a

Claims (2)

An unmanned mobile device made movable;
A sensor that is mounted on the unmanned mobile device and measures the presence or absence of fixed obstacles and moving obstacles around the unmanned mobile device;
The unmanned mobile device is mounted on the unmanned mobile device, detects a movable area without the fixed obstacle based on the measurement result of the sensor, determines a planned movement path of the unmanned mobile device within the movable region, and the unmanned mobile device. The moving device is controlled to move on the planned moving path, and the moving obstacle is detected based on the measurement result of the sensor, and the moving path of the moving obstacle is predicted to collide with the moving obstacle. And a control means for avoiding
In the case where a boundary where the fixed obstacle or the unmeasured area due to the moving obstacle contacts the movable area is found in a direction perpendicular to an arbitrary point on the planned movement path, Except when the distance from the arbitrary point to the boundary is a distance that allows a collision avoidance operation after a moving obstacle has popped out of the unmeasured area, the start position of the search may be popped out. If the moving obstacle is detected by braking so that it can decelerate or stop until the position where the popping possibility section starts from the unmanned moving device, the stopped state is continued and there is no moving obstacle. In this case, the autonomous mobile device returns to the initial speed and resumes the movement along the planned movement route. An autonomous mobile device control method for controlling movement of the unmanned mobile device based on a measurement result by a sensor that measures the presence or absence of a surrounding fixed obstacle and a mobile obstacle mounted on a movable unmanned mobile device,
Detecting a movable area without the fixed obstacle based on the measurement result by the sensor, determining a planned movement path of the unmanned mobile device in the movable area,
Detecting the moving obstacle based on the measurement result by the sensor, predicting the movement path of the moving obstacle,
While controlling the unmanned mobile device, avoiding a collision with the moving obstacle, moving on the planned movement path,
Perpendicularly from an arbitrary point on the moving prediction path, the fixed obstacle, or, when said boundary unmeasured region by moving obstacle and said movable area in contact is found, from the arbitrary point Except when the distance to the boundary is a distance that allows a collision avoidance operation after a moving obstacle has popped out of the unmeasured area, the search start position is set as a popping possibility section, and the unmanned mobile device When braking is detected so that it can be decelerated or stopped until the position where the pop-out possibility section starts, and when a moving obstacle is detected, the stop state is continued. The method for controlling an autonomous mobile device is characterized by returning to the speed of the vehicle and resuming movement along the planned travel route.

Images