Generic UAM Simulation - NASA TTT-Autonomous Systems(AS): Intelligent Contingency Management (ICM)

A MATLAB/Simulink framework that models a NASA Lift + Cruise VTOL.

Quick Start

% From repository root:
>> main           % launches default config params & opens GUAM model

Simulation output is logged to logsout{1} and can be mapped to a structure:

>> SimOut = logsout{1}.Values;

About the Simulation

This simulation is a generic UAM simulation. It includes a generic transition aircraft model representative of a NASA Lift+Cruise vehicle configuration.

Some of the key simulation components include:

1. A simulation architecture (e.g., signal buses) that supports the most common rigid body 6-DOF frames of reference (e.g., Earth Centered Inertial, Earth Centered Earth Fixed (ECEF), North-East-Down (NED), Navigation, Velocity, Wind, Stability, and Body)
2. A simulation architecture that contains most aerospace signals/quantities of typical interest
3. A generic architecture that readily supports swapping in and out aircraft models, sensors, actuator models, control algorithms, etc.

Simulation input (fixed) parameters are provided in a large structure SimIn, whereas desired tunable simulation parameters are provided using the large structure: SimPar. The structure SimIn, SimPar, and SimOut therefore contain the (fixed) simulation inputs, the (variable) simulation inputs, and the simulation outputs respectively. Some basic results plotting can be performed by running the m-file: ./vehicles/Lift+Cruise/Utils/simPlots_GUAM.m. Simulation results animation (e.g., creation of a .avi file or similar) is available by use of the script: ./utilities/Animate_SimOut.m.

Simulation Modifications by ADCL ERAU

This simulation was originally forked from NASA's model. Several modifications were introduced:

• Controller Replacement: The default baseline controller was replaced with an Incremental Nonlinear Dynamic Inversion (INDI) controller, following the architecture proposed by Lombaerts et al. and implemented in our work:
  • Lombaerts et al., AIAA 2020-1619
  • ADCL ERAU Implementation, AIAA 2025-3489
• Remote Control & Visualization Blocks: Simulink blocks were added to allow remote control inputs and visualization of the vehicle state.
• VTOL Scaling Script: A dedicated script was created to scale the VTOL model based on Froude Scaling and Similitude principles. The method follows:
  • Wolowicz, C.; Bowman, J.; Gilbert, W. – NASA TP-1534 (1979)
• Trim Script for Scaled VTOL: An additional script was implemented to compute trim states for the scaled vehicle configurations.

Personalization

While demonstration scripts are provided, users can customize various subsystems through the userStruct structure and simSetup.m script. userStruct allows setting different subsystem variants and switches to tailor simulations to specific needs. Key configuration options include actuator type, atmosphere model, trajectory input, and more.

For example, a typical userStruct setup might look like this:

userStruct.variants:
    refInputType: Timeseries
    vehicleType: LiftPlusCruise
    expType: DEFAULT
    atmosType: US_STD_ATMOS_76
    turbType: None
    ctrlType: BASELINE
    actType: FirstOrder
    propType: None
    fmType: Polynomial
    eomType: Simple
    sensorType: None

For detailed instructions on setting each option, reference input types, customizing the aero-propulsive model, and other aspects, please see the Personalization Guide.

Simulation Trimming

The (offline) trim routines are found in the ./vehicles/Lift+Cruise/Trim folder. The top-level trim routine is: trim_helix.m. This script was used to trim the overactuated Lift+Cruise vehicle using the polynomial aero-propulsive database. NOTE: the routine could also be used for trimming with the strip theory S-function aero-propulsive model, but the code has not been modified to switch between the models (likely not functional using the S-function model). In the top-level trim_helix.m script, the user specifies a range of forward and vertical velocities (could also provide a turn radius). Next, the user provides some quantities needed for the quadratic cost function/optimization (e.g., initial guess, offset, scaling, and free variables). The quadratic cost function used by fmincon is: mycost.m, and the non-linear constraints function is nlinCon_helix.m. The results of the trim table schedule are then saved in a .mat file.

Controller Architecture

The controller implemented in this simulation is based on the Incremental Nonlinear Dynamic Inversion (INDI) framework as described in AIAA 2025-3489. The implementation follows the control architecture originally proposed by Thomas Lombaerts et al. in AIAA 2020-1619, which was developed for unified, full-envelope flight control of eVTOL vehicles.

Failure Configuration

Inject surface jams, engine cut-outs, sensor biases and more using the failure API described in Failures Documentation.

Further Reading

Area	Markdown
Personalisation guide	Documentation/Personalize.md
Failure-injection API	Documentation/Failures.md
Linearisation process	Documentation/Linearization.md
S-Function build	Documentation/SFunction.md
Vehicle scaling	Documentation/Scaling.md
Trim file creation	Documentation/Trim.md
Frames & symbols	Documentation/RefFrames.md

Related Publications

• VTOL Urban Air Mobility Concept Vehicles for Technology Development Introduces NASA's Lift+Cruise concept.
• A Strip Theory Approach to Dynamic Modeling of eVTOL Aircraft Describes the derivation of a strip theory aerodynamic model.
• Full-Envelope Aero-Propulsive Model Identification for Lift+Cruise Aircraft Using Computational Experiments Describes the derivation of a polynomial-based aerodynamic model.