This project report deals with work done for implementing and testing different highlevel con-trollers for path planning algorithms in a simulated en-vironment. The work is done in a simulated environmentand the task of autonomous navigation in an unknown environment is projected. The project is divided intotwo parts , the first step is to navigate in an unknown environment for certain time to create a map of theplaced system. The second step is to return back to the home position from any point the exploration is stopped.
Exploration is carried out as shown in the figure below. Evidence grid is created to store the probability of the corresponding region in any space. It can fuse information from different types of sensors. Cells are initialized at a prior probability withrough estimate of overall probability at any given location. Once evidence grid there, for each cell occupancy probability is compared with the initialprobability and they are classified.
This is exploration carried out for around 50 min in willow garage map.
For returning home, this is the path taken.
The robot was able to successfully explore the environment.