Date of Award

4-2023

Document Type

Dissertation

Publisher

Santa Clara : Santa Clara University, 2023.

Degree Name

Doctor of Philosophy (PhD)

Department

Mechanical Engineering

First Advisor

Christopher Kitts

Abstract

Adaptive navigation is a subcategory of navigation techniques that attempts to identify goal locations that satisfy specific criteria in an unknown area. In 2D scalar field adaptive navigation (SFAN), primitives navigate to or along features of interest in an unknown, possibly time-varying, planar scalar field. Features include extrema, contours, and fronts. This work solves the 2D SFAN problem using swarm robotic techniques. Robotic swarms are a subset of multi-robot systems that use decentralized control of simple interchangeable robots to perform collective actions. A subgroup of swarms is the Reactive Particle Swarm (RPS), characterized based on its simplicity, reactivity to its current environment, and flexibility of applications. Previous work in RPS lacks a unified implementation for RPS behaviors making cross-comparison and reuse challenging.

This work presents a novel 1) RPS control architecture that streamlines the development of novel RPS behaviors, 2) elliptical aggregation algorithm that meets the four tenets of elliptical aggregation, and 3) series of 2D RPS SFAN primitives, and verifies all RPS base and composite behaviors using simulated and hardware-in-the-loop case studies.

The architecture unifies the development of new RPS behaviors. The weighted summation of simple base behaviors and external command inputs form complex composite behaviors. This plug-and-play design concept allows for the rapid development of novel combinations of base behaviors, and emphasizes the topdown design of composite behaviors. A series of simulated and on-hardware case studies demonstrate the utility and flexibility of the architecture while establishing a library of verified RPS base behaviors.

The four tenets of elliptical aggregation are 1) guidelines for swarm and ellipse parameter selection to ensure successful aggregation, 2) commandable ellipse parameters, 3) simplicity for scaling in the number of robots, and 4) adaptive sizing. The elliptical attraction behavior can be leveraged for SFAN to orient the swarm to improve feature sensing and size to overcome noise thresholds. The elliptical attraction behavior and adaptive sizing variant were verified using simulated and experimental trials.

For 2D RPS SFAN primitives, the extremum seeking, contour following, and front identification behaviors and their adaptive sizing variants are verified using simulations incorporating both artificial and interpolated real-world scalar fields and hardware-in-the-loop trials. The ridge descent, trench ascent, and saddle point identification behaviors are presented in a preliminary form and are verified through simulation.

Overall this work has four main contributions, 1) a novel RPS control architecture that unifies the implementation and streamlines the development of novel RPS behaviors, 2) a novel elliptical attraction behavior, 3) novel SFAN primitives, and 4) verification of all RPS behaviors through simulation and hardware-in-theloop trials.



Download

Share

COinS