Path Planning and Obstacle Avoidance Software
Two Projects | Two Approaches
Spring 2018
Two projects are presented here. Both were completed in tandem during construction of the SUNGROVE, with the goal of enabling navigation of an environment using collision avoidance maneuvers and minimum energy consumption. The bicycle model is used to model the behavior of the UGV in both works. Below are the final reports of each study.
​
Gradient Descent
Model Predictive Control (MPC)
A path planning algorithm which uses gradient based optimal control to find the minimum energy solution to the path planning problem with obstacles as constraints is created using a nonlinear bicycle model to model the system intended for use.
A Model Predictive Control (MPC) algorithm for constrained path planning of a UGV. A linearized bicycle model is used to model the systems behavior under a given control input, and an interior point optimization method is used in MATLAB to solve the path planning problems.
Project Requirements
-
Extensive MATLAB usage
-
Mathematics
-
Classical and advanced optimal control theory
-
Excellent written communication
​
​
​