RU2769710C1 - Method for building a route and controlling the movement of a mobile service robot in retail premises - Google Patents

Abstract

FIELD: robotics.

SUBSTANCE: invention relates to the field of robotics and can be used for controlling the movement of a mobile service robot performing process functions at commercial facilities, preventing collisions thereof with obstacles. The method includes defining a route from the route formation module, obtaining a "potential collision" indicator from the collision prevention module, establishing the starting and target points of movement of the robot, and preliminarily indicating the location of obstacles in the map. When an obstacle is detected by sensors in the information processing module, a two-dimensional point cloud of the obstacle is formed and transmitted together with the data on the dimensions of the contour of the platform of the robot to the information processing module, then to the collision prevention module, and then to the movement control module, wherein the linear and angular velocities of the robot are set and transmitted to the module of the electric engine control driver and to the controllers of the electric engines of the wheels.

EFFECT: use of the invention allows for an increase in the accuracy of movement of the robot along a preset trajectory in conditions of a dynamic environment.

1 cl, 5 dwg

Description

The invention relates to the field of robotics, and in particular to methods for constructing a route of movement and avoiding collisions with obstacles of a mobile service robot (hereinafter also referred to as a "robot") that performs technological functions at commercial facilities, in particular, inventory-related operations.

An autonomous mobile robot is known, which includes a robot base, an on-board navigation system and a method for its movement ( US patent No. 9592609 dated January 25, 2013) in a physical environment with stationary and non-stationary obstacles, and the method includes:

a) saving on the basis of the robot a global map that defines the layout of the premises with a set of jobs;

b) receiving by the robot from the base of the job, including the specified location of the job associated with the map with the specified set of operations;

c) activation of the onboard navigation system for automatic

(i) determining the location of the robot based on the global map;

(ii) construction of a local trajectory in a two-coordinate system using sensors and further refinement of the local trajectory;

(iii) finding the shortest path using the local path on the grid map and moving the mobile robot from the current position of the mobile robot to the target point on the floor plan by moving the mobile robot at a given linear and angular velocity along the first path if the mobile robot moves along the first path can lead to a collision with stationary or non-stationary obstacles, then a second path is automatically created from the current position of the mobile robot to the position on the floor plan in accordance with the local trajectory, taking into account collision avoidance;

(iv) issuing motion commands indicating linear and angular velocities and bringing the mobile robot using the second path to the actual location;

d) the execution of the task by the robot as a set of operations associated with the specified location of the task on the map.

The disadvantage of this method is that when building and rebuilding the route of movement, only the start and end points of the route are taken into account, and the intermediate points of the route are not mandatory for passing. Also, when predicting a possible collision of a robot with a stationary or non-stationary obstacle, the physical dimensions of the robot are not taken into account, and when an obstacle is detected by the robot's sensors while the robot is moving, the possibility of a collision with an obstacle after a command to stop the robot is not taken into account.

A device and a method for planning the path of a mobile robot are known ( US patent No. 10369695 of 04/28/2015). The device comprises: an obstacle recognition unit located between the starting point and the target point of the mobile robot for determining nodes based on the grid; a path planning block that selects nodes based on the grid to deploy a path planning tree for motion control of the mobile robot.

The mobile robot path planning method of claim 1, wherein the path planning module performs grid-based node sampling by randomly extracting a random node from a set of nodes in the grid. The mobile robot path planning method of claim 1, wherein the path planning module performs tree expansion planning using the Rapid Random Tree Exploration (RRT) algorithm.

The mobile robot path planning method according to claim 1, wherein the path planning module compares the first function, which includes calculating the distance from the selected to the target node and the direction vector, with the second function, which includes calculating the distance from the current position of the robot to the target point and the direction vector.

The mobile robot path planning method of claim 5, wherein as a result of comparing the first path calculation function with the second path calculation function, the path planning block does not add the selected node to the path planning tree when the value of the first path calculation function is greater than the value of the second calculation function path, and adds the selected node to the path tree when the value of the first path evaluation function is less than the value of the second path evaluation function.

The path planning method of the mobile robot according to claim 1, wherein the recognition unit sets the nodes based on the grid, taking into account the kinematic characteristics of the mobile robot.

The mobile robot path planning method according to claim 8, wherein the path planning unit performs grid-based node sampling without taking into account the kinematic characteristics of the mobile robot.

The method for constructing a mobile robot path includes setting a start point and a target point in the mobile robot motion area based on a capture map, recognizing obstacles located between the start point and the target point of the mobile robot to set up a grid of target points, and selecting nodes based on expanding the path planning tree for control. movement of the mobile robot.

The path planning method of the mobile robot according to claim 11, wherein, when setting the nodes based on the grid, the nodes are set as intermediate points, which are the turning points defined between the start point and the target point based on the map.

The path planning method of the mobile robot according to claim 11, wherein when installing nodes based on a grid, the nodes are set taking into account the kinematic characteristics of the mobile robot.

With this path planning method, the kinematic characteristics of the mobile robot include at least one of the following types of information: information about the movement of the mobile robot or information about the movement of the mobile robot. Path planning tree expansion and grid-based node selection are performed by randomly inserting any node from the grid-based nodes and iterated until the selected node is a node corresponding to a point, possibly using the fast random tree exploration ( RRT ) algorithm.

The disadvantage of this method is that when an obstacle is detected by the robot's sensors while the robot is moving, the possibility of a collision with an obstacle after a command to stop the robot is given, and the absence of a trajectory controller module worsens the accuracy of the robot's movement along a given trajectory.

A system and a method for navigating a robot along a target trajectory with obstacle avoidance are known ( US patent No. 10429847 dated October 1, 2019), in which other moving robots are dynamic obstacles.

A system for navigating a robot along a trajectory to a target with obstacle avoidance provides a transceiver, a processor, and a data storage device for instructions, for the processor to perform calculations, to reach the target, determine paths to the target, obtain an obstacle map, position coordinates of the robot, obtain the positions of one or more others. robots as dynamic obstacles for the robot, generating a set of robot speeds; estimating the first set of robot speeds using the functions; choice of preferred robot speed; creating a set of obstacle velocities.

The method for navigating a robot through trajectory tasks with obstacle avoidance includes: obtaining a target for the robot, determining the trajectory, obtaining an obstacle map, obtaining the position of the robot, generating a set of speed for the robot, estimating, using functions, a set of speeds, choosing a preference for the speed of the robot. Basic functions, definition and characteristics of dynamic obstacles based on the position of the robot and the preferred speed of the robot; move the robot based on the preferred speed.

The method of claim 1, wherein the pose of the first robot is determined by one or more of the following: many-to-many multi-resolution scan matching (M3RSM), Monte Carlo adaptive localization (AMCL), global positioning system (GPS) , reference information and odometry based on robot sensors.

The disadvantage of this method is the reduced accuracy of setting the parameters of the robot's movement along a given trajectory in a dynamic environment.

A known method of mixed path planning is used to control the movement of mobile robots in rooms using a dynamic window and the optimal method of mutual collision avoidance in a dynamic environment ( CN patent No. 109945873 dated 09/22/2017, prototype of the method).

The mixed path planning method is combined with the global path planning method in a static environment and the local path planning method. The mixed path planning method is used to supply information about the external environment to the system through an external sensor to build a grid model. The invention discloses a mixed path planning method combining a parallel bidirectional A-star algorithm and an improved artificial potential field algorithm; a parallel bidirectional A-star algorithm is used to find global and static paths; an improved artificial potential field (APF) algorithm is used to find local and dynamic paths.

The use of the above methods when controlling the movement of the robot involves:

- receiving a route from the software module for forming the route of movement;

- receiving the sign "potential collision" from the collision avoidance software module;

- obtaining coordinates of the location of the robot from the software module for generating a global map of the premises;

- setting the starting point and target point of the robot's movement;

- preliminary indication of the location of obstacles on the map;

- searching and obtaining points of the optimal route of movement without collisions from the current position to the target using the path planning algorithm;

- construction of a local trajectory without collisions, and extraction of a number of local target points, relative to the starting point, turning points and target point;

- building a part of the local trajectory and updating information about the local trajectory using the results of scanning by the robot's sensors in real time;

- launching the process of controlling the movement of the robot during local path planning, reading information from the local trajectory provided by the information flow from the local trajectory building module in real time, using the local target point as the current target point using the path planning algorithm to implement movement along the local trajectory without collisions with a detour of static and dynamic obstacles.

The disadvantage of this method is the reduced accuracy of setting the parameters of the robot's movement along a given trajectory in a dynamic environment.

The technical problem solved by the invention is to improve the accuracy of the movement of a mobile robot along a given trajectory in a dynamic environment.

The technical problem is solved in a way to control the movement of a mobile service robot, consisting of a mobile platform, wheel motors with a controller, sensors, a sensor controller, a human-machine interface unit, a communications unit and a service device, a control system having:

- a computer placed on the mobile platform of the mobile service robot, configured for motion control;

- a software module that controls the mobile platform and provides:

- bypassing obstacles, detecting an obstacle and receiving data from sensors in the form of a two-dimensional cloud of points of detected obstacles and the contour of a mobile service robot, the following actions (calculations) are performed by the collision avoidance software module:

до препятствия в состоянии движения мобильного сервисного робота с учетом предварительной команды управления движением мобильного сервисного робота:- - calculation of the safe distance

to an obstacle in the state of movement of the mobile service robot, taking into account the preliminary command to control the movement of the mobile service robot:

,

,

– безопасное расстояние до препятствия в состоянии движения мобильного сервисного робота,where

– safe distance to the obstacle in the state of movement of the mobile service robot,

– безопасное расстояние до препятствия в состоянии покоя мобильного сервисного робота,

is the safe distance to the obstacle at rest of the mobile service robot,

– линейная скорость мобильного сервисного робота;

is the linear speed of the mobile service robot;

– линейное ускорение мобильного сервисного робота;

– linear acceleration of the mobile service robot;

– ожидаемый тормозной путь мобильного сервисного робота;

is the expected stopping distance of the mobile service robot;

,

,

n-ых экстраполированных позиций мобильного сервисного робота:- calculation of coordinates by the program

,

,

n -th extrapolated positions of the mobile service robot:

,

,

,

,

,

,

,

– координаты экстраполированной позиции мобильного сервисного робота;where

,

– coordinates of the extrapolated position of the mobile service robot;

– угол направления движения мобильного сервисного робота в экстраполированной позиции;

is the angle of the direction of movement of the mobile service robot in the extrapolated position;

– номер экстраполированной позиции мобильного сервисного робота;

is the number of the extrapolated position of the mobile service robot;

– шаг экстраполяции по времени;

is the time step of extrapolation;

- issuing a command to move the projection of the form of the mobile service robot to an extrapolated position;

- subsequent determination of whether the projection of the form of the mobile service robot crosses the two-dimensional cloud of obstacle points.

This set of distinguishing features of the method for controlling the movement of a mobile service robot was not found in the process of patent information search, therefore, the invention meets the criterion of "novelty". It also does not follow explicitly from the prior art, therefore, the invention meets the criterion of "inventive step".

In FIG. 1 shows a diagram of a commercial premises.

In FIG. 2 shows a block diagram of a mobile service robot.

In FIG. 3 is a diagram of the collision avoidance algorithm.

In FIG. 4 shows a diagram of how the anti-collision software module determines the safe distance between the projection of the mobile service robot and the obstacle point cloud.

In FIG. 5 - diagram of the algorithm of the software driver for controlling the wheel motors.

The method of controlling (Fig. 3-5) the movement of a mobile service robot 1 (Fig. 1) in a room 2 of a commercial facility (Fig. 1), consisting of a mobile platform 3, sensors 4, a controller 5 (Fig. 2) of sensors 4, electric motors 6 wheels 7 (Fig. 6), a controller 8 of electric motors 6, wheels 7, a human-machine interface unit 9, a communications unit 10 and a service device 11 using a control system (Fig. 3-4) having:

- a computer 12 placed on the mobile platform 3 of the mobile service robot 1 configured for motion control.

The collision avoidance algorithm (Fig. 3) includes actions (steps) for:

- receipt by the program of a preliminary command to control the movement of the mobile service robot 1 from the program module of the trajectory controller;

- obtaining by the program a two-dimensional cloud of points from the software processing of information from the sensors 4;

до препятствия 23 в состоянии движения мобильного сервисного робота 1 с учетом предварительной команды управления движением мобильного сервисного робота 1:- calculation by the program of a safe distance

to the obstacle 23 in the state of movement of the mobile service robot 1, taking into account the preliminary command to control the movement of the mobile service robot 1:

,

,

– безопасное расстояние до препятствия 23 в состоянии движения мобильного сервисного робота 1,where

– safe distance to the obstacle 23 in the state of movement of the mobile service robot 1,

– безопасное расстояние до препятствия в состоянии покоя мобильного сервисного робота 1,

is the safe distance to the obstacle at rest of the mobile service robot 1,

– ожидаемый тормозной путь мобильного сервисного робота 1;

is the expected stopping distance of the mobile service robot 1;

,

,

n-ых экстраполированных позиций мобильного сервисного робота 1:- calculation of coordinates by the program

,

,

n -th extrapolated positions of the mobile service robot 1:

,

,

,

,

,

,

,

,

– экстраполированные координаты позиции мобильного сервисного робота 1;where

,

,

– extrapolated position coordinates of the mobile service robot 1;

– номер экстраполированной позиции мобильного сервисного робота 1;

– number of extrapolated position of mobile service robot 1;

– шаг экстраполяции во времени;

is the extrapolation step in time;

- moving the projection of the shape of the mobile service robot 1 to an extrapolated position with the subsequent determination of the presence of its intersection with a two-dimensional cloud of points 24 (Fig. 4).

The algorithm (Fig. 5) of the software driver for controlling electric motors 6 of wheels 7 includes:

- receiving by the program a command to control the movement of the mobile service robot 1 from the collision avoidance software module;

и

первого и второго колес 7 робота 1 с радиусом

и колеей

:- calculation by the program of the task for the controller 8 electric motors 6 wheels 7 speeds

and

of the first and second wheels 7 of the robot 1 with a radius

and track

:

,

,

,

,

– скорость первого колеса 7 мобильного сервисного робота 1,where

is the speed of the first wheel 7 of the mobile service robot 1,

– скорость второго колеса 7 мобильного сервисного робота 1,

is the speed of the second wheel 7 of the mobile service robot 1,

– величина колеи мобильного сервисного робота 1,

– gauge of the mobile service robot 1,

– радиус колеса мобильного сервисного робота 1.

is the wheel radius of the mobile service robot 1.

The technical effect - increasing the accuracy of the movement of the mobile service robot 1 along a given trajectory in a dynamic environment - is achieved due to the distinctive features of the method for controlling the mobile service robot 1 using the described control system:

- after the formation of the trajectory of the mobile service robot for each known fragment of the trajectory, the software module of the trajectory controller calculates the linear and angular speeds of the mobile service robot 1, taking into account the predetermined threshold maximum values of the linear and angular velocities of the mobile service robot 1:

,

,

,

,

– линейная скорость мобильного сервисного робота 1,

– угловая скорость мобильного сервисного робота 1,

– пороговая (максимальная) линейная скорость мобильного сервисного робота 1,

– пороговая (максимальная) угловая скорость мобильного сервисного робота 1:where

– linear speed of mobile service robot 1,

is the angular velocity of the mobile service robot 1,

– threshold (maximum) linear speed of the mobile service robot 1,

– threshold (maximum) angular velocity of the mobile service robot 1:

- when an obstacle 23 is detected and data is received from sensors 4 in the form of a two-dimensional point cloud of detected obstacles 23 and the contour of a mobile service robot 1, the collision avoidance software module performs the following actions (calculations):

до препятствия 23 в состоянии движения мобильного сервисного робота 1 с учетом предварительной команды управления движением мобильного сервисного робота 1:- calculation of the safe distance by the program

to the obstacle 23 in the state of movement of the mobile service robot 1, taking into account the preliminary command to control the movement of the mobile service robot 1:

,

,

– безопасное расстояние до препятствия 23 в состоянии движения мобильного сервисного робота 1,where

– safe distance to the obstacle 23 in the state of movement of the mobile service robot 1,

– безопасное расстояние до препятствия 23 в состоянии покоя мобильного сервисного робота 1,

– safe distance to obstacle 23 at rest of mobile service robot 1,

– линейная скорость мобильного сервисного робота 1;

– linear speed of mobile service robot 1;

– линейное ускорение мобильного сервисного робота 1;

– linear acceleration of the mobile service robot 1;

– ожидаемый тормозной путь мобильного сервисного робота 1;

is the expected stopping distance of the mobile service robot 1;

,

,

n-ых экстраполированных позиций мобильного сервисного робота 1:- calculation of coordinates by the program

,

,

n -th extrapolated positions of the mobile service robot 1:

,

,

,

,

,

,

,

– экстраполированные координаты позиции мобильного сервисного робота 1;where

,

– extrapolated position coordinates of the mobile service robot 1;

– угол направления движения мобильного сервисного робота 1 в экстраполированной позиции;

is the angle of the direction of movement of the mobile service robot 1 in the extrapolated position;

– номер экстраполированной позиции мобильного сервисного робота 1;

– number of extrapolated position of mobile service robot 1;

– шаг экстраполяции во времени;

is the extrapolation step in time;

- issuing a command to move the projection of the form of the mobile service robot 1 to the extrapolated position;

- subsequent determination of the presence of the intersection of the projection of the shape of the mobile service robot 1 two-dimensional cloud of points 24 obstacles 23 (marked with a blue figure).

Claims (16)

A method for controlling the movement of a mobile service robot, consisting of a mobile platform, wheel motors with a controller, sensors, a sensor controller, a human-machine interface unit, a communication unit, a service device and a control system containing a computer placed on said mobile platform of the robot and configured to control its movement using: - a software module for generating a static map of the premises of a commercial facility with a division into zones and indicating the order of scanning zones, - navigation software module, - software module for processing information from lidars and ultrasonic sensors, - software module for the formation of the route of movement, - motion control software module, - software driver control module for electric motors, - collision avoidance software module, including: - receiving a given route from the software module for generating a route, - receiving the sign "potential collision" from the collision avoidance software module, - setting the starting point and target point of movement of the mobile service robot and - preliminary indication of the location of obstacles on the map, characterized in that when an obstacle is detected by the sensors in the sensor information processing software module, a two-dimensional obstacle point cloud is formed, which is transmitted together with data on the dimensions of the mobile service robot platform contour to the sensor information processing software module, and then this data is transmitted to the collision avoidance software module and then to the software module. a motion control module in which the task of linear and angular speeds of movement of the mobile service robot is formed and transferred to the module of the software driver for controlling electric motors and to the controllers of electric motors of the wheels.

Images