DIJKSTRA - Path Planning Algorithm for a Point Robot
The task was to designing a point robot that would traverse a map where the obstacels are already known to us.
We have a point robot that can move in eight directions and each action would have its respective costs.
Defining the map
We use the concepts of Algebraic Half planes to define the free space and the obstacles.
Exploring the Map
Use the defined actions set and as per the cost for each step, we traverse the graph
Action Sets = {(1,0), (-1,0), (0,1), (0,-1), (1,1), (-1,1),(1,-1),(-1,-1)}
In each action set generated we check if the new position is going to end up in the obstacle space.
Optimal path
After we explore the entire map, we use backtracking to find the path with the least cost.
The operation of the algorithm is shown below.
Usage
- Clone the repo to your local machine
git clone https://github.com/HemanthJoseph/Dijkstra-Path-Planning-Point-Robot.git
- Change Directory
cd Dijkstra_Algorithm
- Run the python file
python Dijkstra.py
- In the command line enter the inputs values for start and goal coordinates and ensure that the points don't fall in the obstacles as the program will keep prompting you to enter points that aren't in the obstacle space.
Dependencies and libraries
- Python 3
- Matplotlib
- Queue
Video Link
https://drive.google.com/file/d/1XJp1R0PteTrX_R4la_C7ffUnGrDoOqul/view?usp=share_link