Formal verification of octorotor flight envelope using barrier functions and SMT solving

07/01/2021
by   Byron Heersink, et al.
0

This paper introduces an approach for formally verifying the safety of the flight controller of an octorotor platform. Our method involves finding regions of the octorotor's state space that are considered safe, and which can be proven to be invariant with respect to the dynamics. Specifically, exponential barrier functions are used to construct candidate invariant regions near desired commanded states. The proof that these regions are invariant is discovered automatically using the dReal SMT solver, which ensures the accurate command tracking of the octorotor to within a certain margin of error. Rotor failures in which rotor thrusts become stuck at fixed values are considered and accounted for via a pseudo-inverse control allocator. The safety of the control allocator is verified in dReal by checking that the thrusts demanded by the allocator never exceed the capability of the rotors. We apply our approach on a specific octorotor example and verify the desired command tracking properties of the controller under normal conditions and various combinations of rotor failures.

READ FULL TEXT
POST COMMENT

Comments

There are no comments yet.

Authors

page 1

page 2

page 3

page 4

03/10/2020

Synthesis of Control Barrier Functions Using a Supervised Machine Learning Approach

Control barrier functions are mathematical constructs used to guarantee ...
01/20/2020

Extent-Compatible Control Barrier Functions

Safety requirements in dynamical systems are commonly enforced with set ...
01/12/2022

Onboard Safety Guarantees for Racing Drones: High-speed Geofencing with Control Barrier Functions

This paper details the theory and implementation behind practically ensu...
10/30/2020

Multi-Layered Safety for Legged Robots via Control Barrier Functions and Model Predictive Control

The problem of dynamic locomotion over rough terrain requires both accur...
09/21/2020

On using SMT-solvers for Modeling and Verifying Dynamic Network Emulators

A novel model-based approach to verify dynamic networks is proposed; the...
05/15/2020

Enhancing Lattice-based Motion Planning with Introspective Learning and Reasoning

Lattice-based motion planning is a hybrid planning method where a plan m...
04/01/2021

Solver-Aided Constant-Time Circuit Verification

We present Xenon, a solver-aided method for formally verifying that Veri...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.