Model Reference Adaptive Control (MRAC) | Theory, Lyapunov Stability & X-29A Flight Control @LearnGandC
Model Reference Adaptive Control (MRAC) | Theory, Lyapunov Stability & X-29A Flight Control  @LearnGandC
Uploaded October 2026 | Updated November 2026, 4 days ago
How does Model Reference Adaptive Control (MRAC) work, and how can adaptive controllers maintain performance in the presence of model uncertainty?

In this extensive one-hour lesson, I develop the fundamentals of multivariable Model Reference Adaptive Control (MIMO MRAC), including the mathematical derivation of adaptive control laws, Lyapunov stability analysis, and application to the lateral dynamics of the X-29A experimental aircraft.

Rather than beginning with elementary single-input systems, we jump into multivariable adaptive control with integral augmentation. The objective is to provide a solid foundation in MRAC theory while connecting the mathematics to a realistic aerospace control problem.

WHAT YOU'LL LEARN
• Transition from Model Reference Control (MRC) to MRAC
• Plant modeling with uncertainties and extended plant development
• Reference model selection and control objectives
• Direct MRAC control architecture and ideal matching conditions
• Derivation of adaptive control laws and tracking error dynamics
• Lyapunov's direct method and stability analysis
• Why stability of the error dynamics alone is insufficient to establish asymptotic tracking
• Barbalat's lemma and proof of asymptotic tracking
• Development of an LQI-based reference model
• Application of MIMO MRAC to the X-29A lateral dynamics

LQI VS. MRAC — WHY THE COMPARISON?

In a previous lesson, we developed a Linear Quadratic Integrator (LQI) controller for the X-29A lateral dynamics. Here, we use that closed-loop system as the MRAC reference model.

Although LQI provides excellent nominal performance and useful robustness properties, its controller gains remain fixed. MRAC introduces online parameter adaptation, allowing the controller to respond to model uncertainty and work toward recovering the desired reference-model performance.

This comparison demonstrates one of the important motivations for adaptive control: the ability to adjust control parameters as the behavior of the plant deviates from the nominal design.

IMPORTANT LIMITATIONS

This lesson focuses on foundational MRAC theory and a simplified X-29A application. The LQI reference model was developed using rigid-body aircraft dynamics without explicitly accounting for actuator dynamics, sensors, aeroelasticity, or digital processing delays. These effects can significantly constrain achievable closed-loop performance.

Practical adaptive controllers also require consideration of parameter drift, wind-up, chattering, robustness, and other implementation challenges. These topics are important for safe and reliable applications but are beyond the scope of this introductory lesson.

The objective is to understand how MRAC works, how adaptive laws are derived, and how Lyapunov theory supports the analysis of stability and tracking.

REFERENCES
Lavretsky, Eugene, and Kevin A. Wise. Robust and Adaptive Control: With Aerospace Applications. 2nd ed., Springer International Publishing, 2024.

Narendra, Kumpati S., and Anuradha M. Annaswamy. Stable Adaptive Systems. Prentice Hall, 1989.

LEARN MORE & SUPPORT THE CHANNEL

Learn Guidance and Control — Access all my lessons for free:
learngandc.com

Support future lessons and engineering education on Patreon:
patreon.com/c/u86359827

TRAJECTI — GUIDANCE & CONTROL APPLIED TO PERSONAL FINANCE

I've also been developing Trajecti, a personal finance application that applies guidance and control principles to everyday spending decisions.

Trajecti provides adaptive daily spending guidance that responds to actual spending to help users stay on course toward their budget goals.

We're currently seeking Android Alpha testers willing to try the application and provide feedback.

Interested in being a human in the loop? Apply here:

trajecti.com/apply

Thanks for watching!

Ben Dickinson
Learn Guidance and Control

Errata:
13:50 \Delta\theta(t) = \hat{\theta}(t) - \theta

#AdaptiveControl #MRAC #ControlTheory #Lyapunov #AerospaceEngineering
Model Reference Adaptive Control (MRAC) | Theory, Lyapunov Stability & X-29A Flight ControlPitch Autopilot and Tuning- Flight Control Fundamentals - Section 1.2 - Rev 2Time to Go Estimation - Guidance Fundamentals II - Section 1.1Autopilot Introduction - Flight Control Fundamentals Section - 1.1Proportional Navigation Intro Solution Set - Section 2 Module 2 - Guidance FundamentalsGain and Phase Margin from Nyquist and Bode - Classical Feedback Control - Section 2 Module 2State Estimation Explained: From Least Squares to Kalman Filters | Lesson 1.1How to Build an X-15 Aircraft Model from Literature | 6-DOF Flight Simulation Tutorial - Section 3.2Attitude Trajectory Shaping Guidance - Guidance Fundamentals IITrim for Autopilot Development – Flight Control Fundamentals – Section 1.7Kinematics for Simulation of Proportional Navigation Rev 2 - Section 3 Module 2 - GuidanceSideslip Rate Feedback Control for Improved Turn Coordination - Lateral Flight Control - Section 1.3
Ben Dickinson |

Model Reference Adaptive Control (MRAC) | Theory, Lyapunov Stability & X-29A Flight Control

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER