Skip to content

Latest commit

 

History

History
34 lines (29 loc) · 2.28 KB

README.md

File metadata and controls

34 lines (29 loc) · 2.28 KB
Raw

Motion Planning in Dynamic and Hostile Environments

alt text

Contents

  1. Introduction
  2. Setup
  3. Run
  4. Output

Introduction

The folloiwing project was carried out as part of the RBE550 Motion Planning course. The aim of the project is to devise a motion planning algorithm for a robot to autonomously navigate to and from a goal location while avoiding both hostile dynamic agents as well as static obstacles. The motion of the robot, from start to goal and then back to the start is planned using a combination of a sampling based algorithm and an algorithm that works well in dynamic environments. The robot in this problem has to navigate to a goal position and then re-navigate back to its start location, in the presence of both static as well as non-deterministic dynamic agents. Therefore, a hybrid algorithm that fuses Rapidly Exploring Random Tree (RRT) to plan the initial path and an Artificial Potential Field (APF) algorithm to avoid dynamic obstacles, was used to solve the problem. Please go through the Project Report for more details.

Setup

The following libraries need to be installed to run the code:

  • numpy
  • matplotlib
  • collections
  • scipy
  • math
  • random
  • impedance_modeles
  • time
  • progress
  • tasks
  • math

Run

Store all the files within the same folder and run the gradient_interactive.py file.

Output

alt text

Results

alt text We successfully carried out a simulation of the search and retrieve maneuver, which is a crucial application scenario for autonomous mobile robots. This maneuver involves the robot locating a target object, planning a path and avoiding dynamic obstacles while traversing to a designated location. The ability to accurately execute the search and retrieve task is essential for various real-world applications such as warehouse automation, package delivery, and environmental monitoring.