NASA Armstrong Flight Research Center · 2026
Actuator Digital Twin & Flight Controls
I developed a physics-based actuator digital twin for fault isolation and a separate four-actuator HIL flight controller.
Expanded technical stack

Overview
For NASA's High-Performance Spaceflight Computing initiative, our team combined physical-model residuals with CNN-based fault identification and isolation. Separately, an experimental four-actuator hardware-in-the-loop flight-control testbed needed an initial controller to exercise and validate its open-avionics software path.
My contribution
I owned the model-based workstream and built a parameterized MATLAB/Simulink actuator twin spanning permanent-magnet synchronous motor and internal torque dynamics, current and voltage behavior, PWM drive behavior, Park and Clarke transformations, thermal response, measured delay, fault dynamics, closed-loop actuator behavior, and supporting subsystems. I calibrated and validated it through system identification, physical frequency-response testing, MATLAB analysis, and actuator and electronics testing. I separately developed the initial controller for the four-actuator testbed.
Challenge
I needed to generate useful model residuals for a fault-classification architecture while matching the dynamics of real actuator hardware closely enough for validation.
Technical approach
I fit physical frequency-response data with least-squares and nonlinear transfer-function identification, incorporated measured delay and thermal behavior, and compared simulated outputs with commanded hardware tests. In the separate controller effort, I implemented discrete PID loops with anti-windup, band-limited derivative components, discrete transfer functions and integrators, and yaw washout filtering to reject low-frequency yaw while retaining higher-frequency disturbances. The software targeted a real-time C++ and RISC-V execution context.
Result / outcome
On doublet validation inputs, the modeled actuator dynamics matched the physical actuator response with R² greater than 0.99. I delivered the twin and validation workflow for continued fault-protection research under the High-Performance Spaceflight Computing initiative. The separate controller was sufficient to validate the initial open-avionics software pipeline, but physical-aircraft testing would require a tuned or different architecture.
Gallery / media
Selected artifacts



