Downwash-aware Configuration Optimization for Modular Aerial Systems

TL;DR

This work tackles the problem of designing task-specific configurations for modular aerial systems while explicitly accounting for inter-module downwash. It introduces a two-stage pipeline: (1) exhaustive enumeration of non-isomorphic, acyclic topologies at fixed connector angles to manage combinatorial complexity, and (2) nonlinear programming to optimize connector angles and rotor inputs for a given wrench set, with downwash constraints modeled via capsule-based collision checks. The approach is applied across configurations with varying module counts, selecting designs that minimize control effort while satisfying actuation and interference constraints, and is validated in physics-based simulation and a real-world toy experiment. The results demonstrate scalable, physically realizable layouts and provide a framework that can be extended to larger modular systems and aerial manipulation tasks.

Abstract

This work proposes a framework that generates and optimally selects task-specific assembly configurations for a large group of homogeneous modular aerial systems, explicitly enforcing bounds on inter-module downwash. Prior work largely focuses on planar layouts and often ignores aerodynamic interference. In contrast, firstly we enumerate non-isomorphic connection topologies at scale; secondly, we solve a nonlinear program to check feasibility and select the configuration that minimizes control input subject to actuation limits and downwash constraints. We evaluate the framework in physics-based simulation and demonstrate it in real-world experiments.

Figures (11)

• Figure 1: The optimal assembly of $12$ modules, given the target wrench set defined in \ref{['sec:4']}. Each black square represents one module, each blue rectangle represents a rigid connection, each red arrow represents the module frame $z$-axis, and the purple capsules indicate the collision-free downwash volume for each module.
• Figure 2: An illustration of two modules with four connection ports on each frame. Each connector can rotate along the green axis, where the orange screw can fix the connector. The white circles on the connector are permanent magnets.
• Figure 3: An example of a four-module assembly and its matrix representation, where a black square represents a single module, the four rectangles within each square are the connectors. Blue indicates a valid connection, yellow indicates a blocked connection, and gray indicates an available connector.
• Figure 4: All non-isomorphic configurations under the symmetric point group $C_4$ using $5$ modules.
• Figure 5: Comparison of exhaustive and sampling-based enumeration. The sample count is fixed at $N_n = 500$ for all $n$. For $n > 11$, CPU time is shown only for every fifth value of $n$.