Constrained Control Allocation With Continuous-Time Rate Constraints: Three-Dimensional Case

TL;DR

This paper introduces a quadratic programming method for control allocation in over-actuated aircraft that directly enforces continuous-time rate constraints without slack variables, improving actuator performance and safety.

Contribution

It presents a novel convex hull-based approach to handle continuous-time rate constraints directly in control allocation, avoiding slack variables and enhancing performance.

Findings

Achieves feasible control allocation with continuous-time rate constraints.

Demonstrates smoother actuator behavior and better constraint satisfaction.

Improves actuator tracking, reduces overshoot, and enhances safety.

Abstract

This paper presents a novel quadratic programming (QP) approach for constrained control allocation that directly incorporates continuous-time actuator rate constraints without requiring slack variables. Over-actuated aircraft configurations, particularly prevalent in eVTOL and military applications, require control allocation algorithms to distribute commanded control moments among available actuators while respecting position and rate constraints. Existing methods such as direct allocation, pseudo-inverse, cascaded generalized inverse, and exact redistributed pseudo-inverse either cannot handle rate constraints in continuous time or require discretization approaches that compromise performance. Current QP methods that incorporate rate constraints rely on slack variables to ensure feasibility, which prevents full utilization of the attainable moment set and degrades allocation performance. The proposed methodology addresses this limitation by calculating the attainable moment set from both position and rate constraints through convex hull operations, then ensuring feasibility by scaling unattainable commanded moments to the boundary of the attainable moment set while preserving their direction. This approach guarantees the feasibility of the optimization problem without slack variables. The method is validated through simulation on an F-18 fighter aircraft control allocation problem, demonstrating equivalent performance to the established exact redistributed pseudo-inverse method while providing smoother actuator behavior and enhanced constraint satisfaction. Results show that incorporating continuous-time rate constraints leads to improved actuator tracking, reduced overshoot, and more precise adherence to position limits, which is essential for aircraft safety, ride comfort, and actuator longevity.