Cooperative 3D Path Optimization (C3PO) Simulation.

摘要

Given a problem to get a group of unmanned vehicles from a starting location to a goal in 3 dimensions, we were able to do so without a home base for communication. Starting out with full map knowledge, the group would elect a leader, plan a path using Rapidly-Exploring Random Trees (RRTs), and move to the goal using Artificial Potential Field. The simulation was created in the MASON multi-agent simulation framework, and we were able to show that RRTs are a viable solution for path planning in 3 dimensions. We were also able to show how certain input parameters affect the ability to plan paths quickly.