ground_robot

ground_robot

This package uses the distancefield package to support a controller for ground robots. It can be used to control a differential drive or a skid-steering robot. Two types of input controls are available, (i) linear and angular speeds, and (ii) wheels speeds. It can be checked with the differential and skid-steering robots available in the robotsim package.

groundrobot_class (python)

This class is an implementation of controllers for wheeled ground robots. The controllers are designed upon the vector field implemented in the distancefield package (check it here). The groundrobot_class object has a member of the distancefield_class called vec_field_obj.

Currently, the groundrobot_class supports two types of robots, the differential drive, and the skid-steer. See the image below:

The available methods are listed below.

Methods

__init__(self, vr, kf, reverse_direction, flag_follow_obstacle, epsilon, switch_dist, d, move_backwards):

Constructor method. It receives the following parameters:

• vr (float): norm of the velocity of the field
• kf (float): convergence gain of the vector field
• reverse_direction (bool): flag to make the vector field follow the curve in the opposite direction
• flag_follow_obstacle (bool): flag to enable/disable the object follow feature
• epsilon (float): distance that a close obstacle will be followed
• switch_dist (float): distance from which an obstacle starts to be followed
• d (float): displacement of the control point used by the feedback-linearization controller
• move_backwards (bool): flag to set the robot to move backwards

set_state(self, state):

Method to set the state of the ground robot. The argument state is a list with the x position, y position, and yaw angle.

set_closest(self, point):

Method to set the 3D position (point) of the closest point that belongs to an obstacle. The third element can be set 0.

get_vw(self):

Method to get the x linear velocity (v_x) and the z angular velocity (w_z) of the robot. It assumes that the robot is in the state previously defined with the method set_state.

feedback_linearization(self,f):

Method that computes the x linear velocity and the z angular velocity for the robot given a reference velocity f. The velocity f is imposed to a point displaced at a distance d in front of the robot. It assumes that the robot is in the state previously defined with the method set_state.

get_wheels_diferential(self,b):

Method to compute the linear velocities of the right and left wheels of a differential drive robot with parameter b (see image above). It assumes that the robot is in the state previously defined with the method set_state. It returns a list with two elements, the right and left speeds.

get_wheels_skidsteer(self, a, b):

Method to compute the linear velocities of the right and left wheels of a skid-steer robot with parameters a and b (see image above). It assumes that the robot is in the state previously defined with the method set_state. It returns a list with two elements, the right and left speeds.

differential_node.py

This ROS node has an implementation that can be used to control a differential drive robot. Basically, it subscribes to a topic to get the robot's pose, computes the vector field and the associate linear and angular speeds for the robot, and publishes the result in a command topic.

It can be tested with differential_sim node available in the robotsim package. See the package examples for instruction in how to launch this simulation.

Parameters

• vr (float): norm of the velocity of the field

• kf (float): convergence gain of the vector field

• reverse_direction (bool): flag to make the vector field follow the curve in the opposite direction

• d_feedback (float): distance of the control point used in the feedback linearization

• move_backwards (bool): flag used to make the robot move backwards

• flag_follow_obstacle (bool): flag to enable/disable the object follow feature

• epsilon (float): distance that a close obstacle will be followed

• switch_dist (float): distance from which an obstacle starts to be followed

• obstacle_point_topic_name (string): name of the topic in which the closest point of the obstacles are published

• pose_topic_name (string): name of the topic in which the robot's pose is published

• pose_topic_type (string): type of the topic in which the robot's pose is published (TFMessage, Pose, Odometry)

• cmd_vel_topic_name (string): name of the topic in which the node publishes command velocities

• path_topic_name (string): name of the topic in which the path (sequence of points) is published

• path_equation_topic_name (string): name of the topic in which the path (equation) is published

Check these parameters in the file config/differential_params.yaml.

Topics

• path_topic_name (message type: distancefield_msgs/Path): Subscribe to this topic to get a path represented as a sequence of points

• path_equation_topic_name (message type: distancefield_msgs/PathEq): Subscribe to this topic to get a path represented by a parametric equation

• obstacle_point_body_topic_name (message type: std_msgs/Point): Subscribe to this topic to get the closest collidable point written in the body reference frame.

• pose_topic_type (message type: tf2_msgs/TFMessage or geometry_msgs/Pose or nav_msgs/Odometry): Subscribe to this topic to get the robot's pose

• cmd_vel_topic_name (message type: geometry_msgs/Twist): Topic in which the velocity command is published (linear velocity)

skidsteer_node.py

This ROS node has an implementation that can be used to control a skid-steering robot. Basically, it subscribes to a topic to get the robot's pose, computes the vector field, and publishes the right and left wheels speeds in a command topic. The difference between this node and the differential_node is that, now, the node actuates directly on the speed of the wheels.

It can be tested with skidsteer_sim node available in the robotsim package. See the package examples for instruction in how to launch this simulation.

Parameters

The parameters of the skidsteer_node are the same as the parameters of the differential_node above with the addition of the following two:

• a (float): horizontal distance from the robot's center to the front (or back) wheels' shafts
• b (float): horizontal distance from the robot's center to the right (or left) wheels' shafts

Check these parameters in the file config/skidsteer_params.yaml.

Topics

The topics of the skidsteer_node are almost the same as the topics of the differential_node above. The only change is the replacement of the topic cmd_vel_topic_name by the topic cmd_wheels_topic_name, which has the following structure:

• cmd_wheels_topic_name (message type: std_msgs/Float32MultiArray): Topic in which the command speeds for the four wheels are published. The array has the order: front right, back right, back left, front left