Task02a - Reactive obstacle avoidance

The main task is to implement a function that will steer the robot towards a given goal while reactively avoiding obstacles.

Deadline 20. October 2018, 23:59 PST
Points 3
Label in BRUTE Task02a
Files to submit archive with Robot.py file and optionally RobotConst.py file
Resources Task02a resource package
Blocks V-REP scene

Assignment

In class Robot.py implement the goto_reactive(coord) function. The purpose of the function is to navigate the robot towards the goal given by coordinates $coord = (x_{goal},y_{goal})$ while reactively avoiding collisions. The steering of the locomotion is achieved by the goto_reactive function by setting the differential steering command of the CPG locomotion controller $(v_{left}, v_{right})$. The function returns True when the robot is at the goal coordinates and False if it has collided with an obstacle en route. In addition to the Task01 - Open-loop locomotion control the function goto_reactiveimplements the reactive collision avoidance.

Information about the current position $(x,y)$, orientation $\phi$ and collision state is provided by the RobotHAL interface through the self.robot object. The respective functions are

#get position of the robot as a tuple (float, float)
self.robot.get_robot_position()
#get orientation of the robot as float
self.robot.get_robot_orientation()
#get collision state of the robot as bool
self.robot.get_robot_collision()

The obstacle sensing is achieved using the simulated laser range finder through the RobotHAL interface through the self.robot object.

scan_x, scan_y = self.robot.get_laser_scan()

The goto_reactive function has a following prescription

def goto_reactive(self, coord):
    """
    Navigate the robot towards the target with reactive obstacle avoidance 
 
    Parameters
    ----------
    coord: (float, float)
        coordinates of the robot goal 
 
    Returns
    -------
    bool
        True if the destination has been reached, False if the robot has collided
    """

Approach

The recommended approach for the reactive obstacle avoidance uses simple AI cognitive model of Braitenberg vehicles described in Lab02 - Exteroceptive sensing, Mapping and Reactive-based Obstacle Avoidance.

The direct sensory-motor mapping can be achieved using the following continuous navigation function (pseudocode).

while not goal_reached:
    dphi = the difference between the current heading and the heading towards the target
    scan_left = closest obstacle to the left of the robot
    scan_right = closest obstacle to the right of the robot
    v_left = 1/scan_left*C_AVOID_SPEED - dphi*C_TURNING_SPEED + BASE_SPEED
    v_right = 1/scan_right*C_AVOID_SPEED + dphi*C_TURNING_SPEED + BASE_SPEED
Where C_AVOID_SPEED is a constant that defines the robot fear of the obstacle, C_TURNING_SPEED is a constant that defines the aggression with which the robot will turn towards the desired heading and BASE_SPEED is the default speed of the robot when it is heading directly towards the target.

Note, that in a physical world it is impossible to get to a precise specific coordinates, therefore it is sufficient to navigate “close enough”. The sufficient distance should be comparable in size to the actual robot. In our case, this distance is given as the navigation parameter DISTANCE_THLD = 0.1 #m defined in the RobotConst.py file.

Evaluation

The code can be evaluated using the following script

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
 
import sys
import math
import time
import numpy as np
sys.path.append('robot')
 
import Robot as rob
 
DISTANCE_THLD = 0.15 #m
 
def check(pose_set, pose_real):
    """
    Function to check that the navigation towards the goal has been successfull
 
    Parameters
    ----------
    pose_set: (float, float, float)
        desired coordinates to reach (x,y,phi)
    pose_real: (float, float, float)
        real coordinates of the robot (x,y,phi)
 
    Returns
    -------
    bool
        True if the robot real position is in delta neighborhood of the desired position, False otherwise
    """
    ret = True
    (x1, y1, phi1) = pose_set
    (x2, y2, phi2) = pose_real
    dist = (x1 - x2)**2 + (y1-y2)**2
    #check the distance to the target
    if math.sqrt(dist) > DISTANCE_THLD:
        ret = False
    #check the final heading
    if not phi1 == None:
        dphi = phi1 - phi2
        dphi = (dphi + math.pi) % (2*math.pi) - math.pi
        if dphi > ORIENTATION_THLD:
            ret = False
 
    return ret
 
if __name__=="__main__":
    #navigation points
    route = [(1.5, 1.5, None), (5, 5, None)]
 
    #instantiate the robot
    robot = rob.Robot()
 
    #navigate
    for waypoint in route:
        pos_des = waypoint[0:2]
 
        print("Navigation point " + str(pos_des))
 
        #navigate the robot towards the target
        status1 = robot.goto_reactive(pos_des)
 
        #get robot real position
        pose_real = robot.get_pose()
 
        #check that the robot reach the destination
        status2 = check(waypoint, pose_real)
 
        #print the result
        print(status1, status2)

The expected output on blocks.ttt map is

Connected to remote API server
Robot ready 
Navigation point (1.5, 1.5) None
True True
Navigation point (5, 5) None
True True

courses/b4m36uir/hw/task02a.txt · Last modified: 2018/10/19 13:11 by cizekpe6