This repository contains scripts for basic planning and navigation algorithms implemented using Python. Each directory has a report that details the implementation of the algorithm.
The robot has to navigate a two dimensional space, avoiding known locations with obstacles, traveling from its initial location to a goal location. Implement the RRT path planning algorithm for two cases, (1) Holonomic Robot and (2) Non-Holonomic Robot. This assumes that localization information i.e., robot’s initial position, obstacle location, goal location are given.
Model Predictive Control deals with determines a trajectory that a robot must follow in order to reach its goal and avoid obstacles. This is formulated as an optimization problem wherein we optimize for the velocities at each time step.
Collision Cone formulation is a planning algorithm that transforms dynamic obstacles to the robot’s perspective and avoids them. This assumes that the location and velocity of the obstacles are known.
- Rahul Sajnani