top of page

CONTROL RESEARCH

Academic research project
Completion as employed researcher
Technion — Israel Institute of Technology
TASKS
Simulation infrastructure
Guidance and control design
Procedural terrain generation
Statistical analysis
GUI for simulation and analysis tools
ADVISOR
Assoc. Prof. Moshe Idan
STATUS
Completed in 2020

While studying for my BSc degree in Aerospace Engineering at the Technion, Israel, I conducted a year-long academic research project aiming to autonomously maintain a drone's height above ground using only laser rangefinders (LRFs). This project continued past my graduation, as a paid researcher, up to its completion with good results.

 

The goal of this project was not only to maintain a given height above ground while also navigating horizontally along a specified path, but to do so with very limited data; each LRF can only provide the position of a single point on the ground relative to the drone at any given time. Using a small number of LRFs allows only a very partial estimation of the ground, with significant inherent errors.

During the course of the research, I defined and simulated the drone's dynamic model, designed controllers to stabilize and command it, incorporated custom Perlin noise terrain generation, implemented raycasting against it to simulate LRF measurements, integrated a guidance algorithm relying on these measurements, and presented a statistical analysis of the system's performance. This simulation and analysis infrastructure was wrapped in graphical applications for ease of use, and a simulation animation tool was built as a side project.

Example simulation results. The black dashed line marks the desired path, curving horizontally along a preset 2D path; and vertically — to match the topography of the ground below. The magenta dotted line denotes the drone's current estimation of the desired path — the preset 2D path projected onto the ground model, calculated in real time using the LRF measurements and offset to the desired height. A portion of this model is visualized by the translucent grid, which can be seen approximating the ground in the LRF measurement points' vicinity.

bottom of page