JP2012128781A - Moving body - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform self-position identification while reducing operation costs.SOLUTION: A moving body 10 includes self-position-identification means 120, environment measurement means 110 of measuring a circumferential environment, obstacle map generation means 130, route planning means 140 of generating a moving route to a destination, tracking control means 150 of generating a control signal for traveling the route, driving means 160 of driving based upon the control signal, and acceleration measurement means 190 of measuring acceleration of the moving body. The self-position-identification means 120 identifies the self position of the moving body by matching with a map when the moving body passes the position of an obstacle recorded in an obstacle map, when acceleration exceeds a threshold, when the moving body passes a place where curvature of the route is equal to or larger than a threshold, when it is determined that a slip is made by comparing the control signal with a driving state of the driving means, when a predetermined time passes after the identification of the self position of the moving body by the matching with the map, or when the moving body travels by a predetermined distance.

Description

  The present invention relates to a moving body.

  In recent years, autonomous mobile bodies that move along a predetermined route have been provided. Such a moving body travels while identifying its own position on the route.

  The self-position estimation of the moving object is based on the combination of the method (odometry) by estimating the self-position based on the driving wheel diameter and the rotation speed of the driving wheel and the method of identifying the self-position by matching with the map. Methods for identifying are known.

  In the self-position estimation by odometry, the self-position is estimated by calculating the moving distance from a certain point by integrating the rotation speed of the drive wheels. Therefore, in the self-position estimation by odometry, an error may occur due to the slip of the driving wheel. Further, errors are accumulated when the moving body travels a long distance.

  In the self-location identification method by matching with a map, the surrounding environment is measured using an external sensor, and the self-location is identified by matching with a map held in advance. Thus, the self-position estimation method based on matching with the map can identify the absolute position of the moving object, but has a feature that the calculation cost is high.

Therefore, in a moving body, self-position estimation by odometry is executed with high frequency, and self-position identification by matching with a map is executed with low frequency. As a result, the mobile body identifies the self-position while correcting the error caused by the self-position estimation by odometry by matching with the map.
Thereby, the moving body can reduce calculation cost compared with the case where only the self-position identification by matching with a map is used.

  Patent Document 1 describes a moving body that can reduce the processing cost of landmark information necessary for identifying its own position and can flexibly cope with changes in the environment.

  Patent Document 2 describes a moving body that corrects a shift in odometry. The moving body stores vehicle feature values for each section in which the vehicle travels, and estimates the self-position based on the vehicle feature values set for each section and information based on the odometry when self-position estimation is performed by odometry. To do.

JP 2008-165275 A JP 2006-252350 A

In self-position estimation by odometry, the timing at which an error occurs is unknown. Therefore, the moving body corrects the error caused by odometry by performing self-position identification by matching with a map at regular intervals. However, since the self-position identification by matching with a map is performed even when correction of an error caused by odometry is not necessary, the calculation cost is high and other tasks may be restricted.
The present invention has been made to solve such problems, and an object of the present invention is to perform self-position identification while reducing the calculation cost.

The mobile body according to the present invention is a mobile body that autonomously moves by estimating self-position by odometry and identifying the self-position by matching with a map, and measures the surrounding environment by self-position identification means. An obstacle map for generating an obstacle map in which the position of the obstacle is recorded based on the environment measurement means for performing the measurement, the information measured by the environment measurement means, and the position information obtained by the self-position identification means Generating means, route planning means for generating a travel route to the destination on the obstacle map generated by the obstacle map generating means, and a control signal for traveling the route planned by the route planning means Tracking control means, driving means for driving drive wheels based on a control signal generated by the tracking control means, and acceleration measuring means for measuring acceleration of the moving body, The self-location identifying means is generated by the path planning means when the obstacle position recorded in the obstacle map passes, when the acceleration obtained by the acceleration measuring means exceeds a predetermined threshold value, and By passing through a portion where the curvature is equal to or greater than a threshold value in the route, by comparing the control signal of the follow-up control means with the drive state of the drive means, and by matching the map When the specified time has passed since the identification of the self-location, or when the vehicle has traveled a predetermined distance from the identification of the self-location by matching with the map, the self-location is identified by matching with the map I do.
Thereby, the mobile body can identify its own position by matching with a map when passing through a place where odometry is likely to shift.

  Self-position identification can be performed while reducing the calculation cost.

1 is a block diagram of a moving body according to a first exemplary embodiment. 3 is an obstacle map according to the first embodiment. It is a figure of the obstruction concerning Embodiment 1. 3 is a flowchart of processing according to the first exemplary embodiment; It is a figure in case the acceleration concerning Embodiment 1 becomes more than a threshold value. It is a figure which shows the matching start timing with the map after the acceleration determination concerning Embodiment 1. FIG.

Embodiment 1
Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of the moving body 10.

The moving body 10 includes an environment measurement unit 110, a self-position identification unit 120, an obstacle map generation unit 130, a route plan unit 140, a follow-up control unit 150, a drive unit 160, a map storage unit 180, an acceleration. Measuring means 190 and drive wheels 170 are provided.
The moving body 10 is a robot that performs autonomous movement. Below, the mobile body 10 may be called a robot.

The environment measuring unit 110 continuously measures the environment around the moving body 10. The environment measuring unit 110 is a distance sensor, for example, a laser distance meter or a stereo camera. The environment measuring unit 110 acquires a distance from a surrounding object as distance data.
The environment measurement unit 110 outputs the acquired distance data to the self-position identification unit 120 and the obstacle map generation unit 130. For example, the environment measuring unit 110 outputs a distance from a surrounding object necessary for identifying the self position to the self position identifying unit 120, and even an obstacle not described in the map stored in the map storage unit 180. Is output to the obstacle map generation means 130.

The self-position identification means 120 performs self-position estimation by odometry and identification of the self-position by matching with a map. The self-position identification unit 120 includes a calculation unit 121 and a storage unit 122.
The self-position identification unit 120 is supplied with map information from the map storage unit 180, is supplied with distance data from the environment measurement unit 110, is supplied with route data from the route planning unit 140, is supplied with encoder values from the drive unit 160, Robot acceleration is supplied from the acceleration measuring means 190. Based on the supplied data and values, the self-position identification unit 120 selects and executes either self-position estimation by odometry or self-position identification by matching with a map. The procedure by which the self-position identifying unit 120 selects a method for obtaining the self-position will be described in detail later.

The calculation unit 121 includes a CPU (Central Processing Unit) that executes processing based on a control program.
When performing self-position estimation by odometry, the calculation unit 121 performs calculation based on the self-position obtained by the previous calculation stored in the storage unit 122 and the encoder value supplied from the driving unit 160. Estimate the current self position. Specifically, the calculation unit 121 calculates that the moving body 10 has moved by a distance based on the encoder value from the self-position obtained by the previous calculation.
When identifying the self position by matching with the map, the calculation unit 121 calculates the self position based on the map information supplied from the map storage unit 180 and the distance data supplied from the environment measurement unit 110. Identify. Specifically, the computing unit 121 computes the position on the map where the moving body 10 exists from the distance data supplied from the environment measuring unit 110. The calculation unit 121 outputs the self-position estimated by the calculation to the storage unit 122 and the obstacle map generation unit 130.

  The storage unit 122 is a storage medium such as a memory or an HDD. The storage unit 122 temporarily stores the calculation result supplied from the calculation unit 121. When the calculation unit 121 estimates the self position by odometry, the storage unit 122 outputs the stored self position at the previous calculation to the calculation unit 121.

The obstacle map generation unit 130 generates an obstacle map based on the obstacle information supplied from the environment measurement unit 110 and the self-position information supplied from the self-position identification unit 120. FIG. 2 shows an example of the obstacle map. In the obstacle map, the location of the moving body 10 on the map and the coordinates of the obstacle are recorded.
FIG. 3 shows a step provided at the entrance of each room as an example of an obstacle. Typically, the obstacle includes an elevator groove and a Braille block in addition to the steps shown in FIG.

Typically, the obstacle map generation unit 130 uses a grid map and a topology map as the obstacle map.
A grid map is a map in which areas on the map are divided in units of grids for each fixed area. In the grid map, the hallway grid, the room A grid, the room B grid, and the room C grid are separated from each other as shown in FIG. The obstacle map generation unit 130 assigns an ID indicating that there is an obstacle to a grid with an obstacle.
The topology map is a map composed of a plurality of nodes and an edge connecting the plurality of nodes. Here, the topology map indicates the relationship between each grid in the grid map. For example, when going from room A to room C, it goes through the entrance of room A, the hallway, and the entrance of room C. In this case, the room A, the entrance of the room A, the hallway, the entrance of the room C, and the room C are set as nodes, respectively, and the entrance of the room A and the room A, the entrance and the hallway of the room A, the entrance of the hallway and the room C, and the room C The entrance and the room C are connected by edges.
In this way, the obstacle map generation unit 130 generates a topology map indicating the relationship between the grids. At this time, the obstacle map generation means 130 sets the location of the obstacle through which the moving body 10 passes as a node.
The obstacle map generation unit 130 sequentially generates an obstacle map and outputs the generated obstacle map to the route planning unit 140. Further, the obstacle map generation unit 130 outputs grid map ID information and topology map node information to the self-position identification unit 120.

The route planning unit 140 plans a moving route along which the moving body 10 moves. The broken line in FIG. 2 is a route planned by the route planning unit 140 for the moving body 10 to move from the hallway to the room B. The route planning unit 140 receives the obstacle map generated by the obstacle map generating unit 130 and plans a route that moves from the current location to the destination. That is, the route planning unit 140 plans a route that travels from the start node to the end node based on the topology map generated by the obstacle map generation unit 130.
The route planning unit 140 outputs the planned route information to the self-position identifying unit 120. Further, the route planning unit 140 outputs route information to the follow-up control unit 150.

  The follow-up control unit 150 controls the moving body 10 to travel along a predetermined route. More specifically, the tracking control unit 150 generates a control signal for controlling the driving unit 160 so that the moving body 10 travels following the route planned by the route planning unit 140. The tracking control unit 150 outputs the generated control signal to the driving unit 160. Further, the follow-up control unit 150 outputs the control value of the generated control signal to the self-position identification unit 120.

The drive unit 160 operates based on the control signal supplied from the follow-up control unit 150 and drives the drive wheels 170. For example, the driving means 160 is a plurality of motors, and drives the driving wheels 170 provided on the left and right of the moving body 10 respectively.
The driving unit 160 includes an encoder 161. The encoder 161 measures the number of rotations of the motor, and outputs the measurement result to the self-position identification unit 120 as an encoder value.

The drive wheels 170 are a plurality of wheels provided on the moving body 10. For example, the drive wheels 170 are provided in a grounded state before and after the moving body 10 and on the left and right.
The driving wheels 170 provided on the left and right of the moving body 10 are respectively connected to the driving means 160, and a driving force is applied from the driving means 160, respectively. Drive wheels 170 provided before and after the moving body 10 are auxiliary wheels to which the drive means 160 is not connected. When the driving force applied from the driving unit 160 is different, the moving body 10 can perform a turning operation because the rotation speeds of the left and right driving wheels 170 are different.
Note that the number and arrangement of the drive wheels 170 are not limited to the above.

  The map storage unit 180 stores map information of an area where the moving body 10 moves. The map storage unit 180 outputs map information to the self-position identification unit 120.

The acceleration measuring unit 190 acquires the acceleration at which the moving body 10 travels. Typically, the acceleration measuring means 190 is an accelerometer. The acceleration measuring unit 190 acquires acceleration in the three axial directions of the moving body 10 including the front-rear direction (x-axis), the left-right direction (y-axis), and the up-down direction (z-axis).
The acceleration measuring unit 190 outputs the acquired acceleration to the self-position identifying unit 120.

  Next, the operation will be described. FIG. 4 is a flowchart in which the self-position identifying unit 120 selects whether to perform self-position estimation by odometry or to identify self-position by matching with a map, and obtain the self-position.

  The self-position identifying unit 120 acquires self-position information (ST10). For example, the self-position identification unit 120 acquires the self-position obtained by the previous calculation, which is stored in the storage unit 122.

  An obstacle map is generated (ST20). Specifically, the environment measurement unit 110 transmits distance data and information on the obstacle to the self-position identification unit 120 and the obstacle map generation unit 130. The obstacle map generation unit 130 generates an obstacle map based on the obstacle information supplied from the environment measurement unit 110 and the self-position information calculated by the self-position identification unit 120. Here, the obstacle map generation unit 130 generates a grid map in which an ID indicating that there is an obstacle is assigned to the area where the obstacle exists and a topology map in which locations where the obstacle exists are registered as nodes.

It is determined whether or not the current position of the moving body 10 is a position where odometry is likely to shift (ST30). The place where the odometry is likely to shift is a place where there is an obstacle, such as a step at the entrance of the room shown in FIG. 3, a place where there is an elevator groove, or a braille block.
More specifically, the self-position identifying unit 120 determines whether or not the position of the moving body 10 is a grid to which an ID indicating that there is an obstacle is assigned in the grid map. In addition, when the mobile object 10 is on the grid to which an ID indicating that there is an obstacle is allocated, the self-position identifying unit 120 determines whether or not the node indicating the obstacle in the topology map has passed.
If the position of the moving body 10 is not in the grid with an ID indicating that there is an obstacle in the grid map, or if it does not pass through the node indicating the obstacle in the topology map (NO in ST30) ), Go to ST40. If the grid map is in a grid with an ID indicating that there is an obstacle and the topology map passes a node indicating an obstacle (YES in ST30), identification of the self-position by matching with the map (ST200).

  The self-position identification unit 120 acquires route information (ST40). Specifically, the route planning unit 140 plans a route from the current location to the target position using the obstacle map generated by the obstacle map generating unit 130. The route planning unit 140 outputs route information to the self-position identifying unit 120.

The self-position identification unit 120 determines whether or not it has passed a location where odometry is likely to shift on the route (ST50). Here, the place where the odometry is likely to shift is a place where the curvature r of the path is equal to or greater than a certain value R. Since the constant value R differs depending on the robot, it is desirable to obtain it by experiment.
If the moving body 10 does not pass through a location where the curvature r is equal to or greater than the predetermined value R (NO in ST50), the process proceeds to ST60. When the mobile body 10 has passed a path where odometry is likely to shift (YES in ST50), the self-position identifying means 120 starts self-position identification by matching with a map (ST200).

  Next, the acceleration measurement means 190 acquires the acceleration of the moving mobile body 10 (ST60). Specifically, vibration and inertia of the moving body 10 are measured using the acceleration measuring unit 190. The acceleration measuring unit 190 outputs the measured acceleration to the self-position identifying unit 120.

The self-position identification unit 120 determines whether or not the acceleration exceeds the threshold (ST70). Here, the situation where the acceleration is equal to or greater than the threshold is as follows. When passing through a level difference on the road surface (Fig. 5 (A)), when passing through an easily slipping environment such as an uneven road surface (Fig. 5 (B)), or in contact with an obstacle such as a wall or a person This is the case (FIG. 5C).
More specifically, the acceleration measuring unit 190 acquires accelerations with respect to the three axes of the x axis, the y axis, and the z axis. The self-position identification unit 120 determines whether or not an acceleration greater than a certain value has occurred. Since the relationship between the odometry shift and the acceleration of each axis and the threshold value to be set differ depending on the robot, it is desirable to obtain them by experiments.
If the acceleration is equal to or less than the threshold value (NO in ST70), the process proceeds to ST80. If the acceleration is equal to or greater than the threshold (YES in ST70), identification of the self-position by matching with the map is started (ST200).

  Here, FIG. 6 is an example of timing for starting matching with a map by detecting a change in acceleration. That is, when it is detected that the acceleration is equal to or greater than a certain value, and the state that is equal to or smaller than the certain value continues for Δt time, the self-position identifying unit 120 starts self-position identification (ST200). Thereby, the vibration and shaking of the moving body 10 are settled and the self-position can be identified after being stabilized.

Next, the rotational speed of the motor provided in the driving means 160 is acquired (ST80). Here, the driving unit 160 operates based on the control instruction given from the follow-up control unit 150 so as to travel along the route planned by the route planning unit 140. For example, the follow-up control unit 150 issues an instruction for the torque to be generated to the driving unit 160.
The encoder 161 provided in the driving unit 160 measures the number of rotations of the motor and outputs it to the self-position identifying unit 120 as an encoder value.

The self-position identification unit 120 determines whether or not the moving body 10 has slipped from the encoder value (ST90). Specifically, the self-position identification unit 120 determines whether the driving wheel is driven when a deviation greater than a certain deviation occurs between the control value output from the follow-up control unit 150 and the encoder value related to the motor rotation speed. It is determined that 170 has slipped. Typically, the slip of the drive wheel 170 occurs when the moving body 10 passes through a road surface with low friction.
If it is determined that drive wheel 170 has slipped (YES in ST90), self-position identification means 120 starts identifying the self-position by matching with the map (ST200). If it is not determined that the vehicle has slipped (NO in ST90), the process proceeds to ST100.

  Next, the processing time from the time when the self-position identification by matching with the previous map is performed is measured (ST100). For example, the self-position identification unit 120 includes a time acquisition unit, and stores the time in the storage unit 122 when the self-position is identified by matching with a map. Further, the self-position identification unit 120 acquires the current time, and acquires the elapsed time from the time when matching with the map was performed last time.

Next, self-position identification means 120 determines whether or not a certain time has passed (ST110). If it is determined that a predetermined time has elapsed (YES in ST110), the self-position identification means 120 is considered to have accumulated errors due to odometry shifts. Is started (ST200).
If the predetermined time has not elapsed (NO in ST110), the process proceeds to ST120.

  The self-position identification unit 120 acquires a travel distance (ST120). For example, the self-position identification unit 120 accumulates the travel distance from the time when the self-position identification was last performed by matching with the map based on the encoder value input from the driving unit 160, and stores it in the storage unit 122. Remember.

  The self-position identification unit 120 determines whether or not the travel distance is equal to or greater than a certain distance (ST130). If it is determined that the vehicle has traveled a predetermined distance (YES in ST130), the self-position identification means 120 is considered to have accumulated errors due to odometry shifts. Is started (ST200). If the travel distance has not reached a certain distance (NO in ST130), the process proceeds to ST140.

  Self-position identification means 120 estimates self-position by odometry (ST140). For example, the calculation unit 121 performs the calculation on the assumption that the vehicle has traveled by the encoder value acquired from the driving unit 160 from the previous self-position. Thereby, the self-position identification means 120 estimates and sets the self-position by odometry. When the self position is set, the process returns to ST10 and the process is repeated until the target position is reached.

  The self-position identification means 120 identifies the self-position by matching with a map when passing through a location where the self-position is likely to shift due to odometry (ST200). More specifically, the self-position identification unit 120 identifies the self-position based on the environment information supplied from the environment measurement unit 110 and the map stored in the map storage unit 180. When the self position is identified, the process returns to ST10 and the process is repeated until the target position is reached.

Thereby, in the mobile body 10, the self-position can be identified by matching with the map when the self-position is frequently estimated by the self-position estimation by the odometry and the portion where the odometry is likely to shift is passed. Therefore, in the self-position estimation by odometry, it deviates from the true value due to the influence of sensor noise, wheel slip, etc., but the deviation can be corrected by identifying the self-position by matching with the map.
Since the self-position is identified by matching with the map when passing through a place where odometry is likely to shift, it is not necessary to perform matching with the map at high frequency, and the calculation cost of the mobile object 10 can be reduced.

Note that the present invention is not limited to the above-described embodiment, and can be changed as appropriate without departing from the spirit of the present invention.
For example, in the above description, the grid map is a single grid in the units of the hallway and the room, but may be any other unit. In addition, both the grid map and the topology map are used as the obstacle map, and when the obstacle node in the topology map is passed, self-position identification is performed by matching with the map. It may be determined whether or not self-position identification is performed by matching with a map by determining whether or not the grid map includes an obstacle. Alternatively, only the passage of an obstacle node in the topology map may be determined without using the grid map information.
The encoder 161 acquires the rotational speed of the motor provided as the driving means. However, the encoder 161 may acquire the rotational speed of the driving wheels 170, for example.

DESCRIPTION OF SYMBOLS 10 Mobile body 110 Environmental measurement means 120 Self-position identification means 121 Operation part 122 Storage part 130 Obstacle map generation means 140 Path planning means 150 Follow-up control means 160 Drive means 161 Encoder 170 Drive wheel 180 Map storage part 190 Acceleration measurement means

Claims (1)

A mobile object that autonomously moves by performing self-position estimation by odometry and identification of the self-position by matching with a map,
Self-localization means;
Environmental measuring means for measuring the surrounding environment,
Obstacle map generating means for generating an obstacle map in which the position of the obstacle is recorded based on the information measured by the environment measuring means and the position information obtained by the self-position identifying means;
On the obstacle map generated by the obstacle map generating means, a route planning means for generating a movement route to the destination;
Follow-up control means for generating a control signal for traveling the route planned by the route planning means;
Drive means for driving the drive wheels based on the control signal generated by the tracking control means;
Acceleration measuring means for measuring the acceleration of the moving body,
The self-position identifying means is generated by the path planning means when the obstacle position recorded in the obstacle map passes, when the acceleration obtained by the acceleration measuring means exceeds a predetermined threshold value, and Matching a map with a case where a path having a curvature equal to or greater than a threshold is passed, a case where it is determined that a slip is generated by comparing the control signal of the tracking control unit and the driving state of the driving unit, and a map When the specified time has passed since the self-location identification by the vehicle, or when the vehicle has traveled a predetermined distance from the self-location identification by the matching with the map, it corresponds to the self-location by matching with the map A moving object that performs identification.

Images