Session: 08-07-02: Novel Control of Dynamic System and Design II

Paper Number: 173714

Enhancing Robotic Trajectory Tracking via Iterative Learning Control and Disturbance Compensation Mechanisms: A Uav Case Study 

Robotic systems operating in dynamic and uncertain environments often face performance degradation due to external disturbances and sensor inaccuracies. From industrial manipulators to aerial platforms, achieving precise and repeatable control is critical for real-world deployment. This talk presents advanced control techniques designed to improve the robustness and accuracy of robotic trajectory tracking, focusing on two complementary approaches: Iterative Learning Control (ILC) and disturbance compensation using Extended State Observers (ESOs).

ILC is a data-driven control methodology particularly well-suited for systems that perform the same task repeatedly. Rather than relying on a detailed plant model, ILC refines the control input over multiple iterations by learning from the tracking error of previous executions. With each repetition, the system improves its performance by adjusting the input signal to reduce the deviation between the actual and desired trajectories. This makes ILC highly valuable in applications such as repetitive path-following, robotic assembly, and aerial maneuvering tasks.

On the other hand, Extended State Observers (ESOs) offer a real-time approach to handling model uncertainties and unknown disturbances. ESOs augment the system state with an additional variable that represents the total disturbance acting on the system—including external forces and internal modeling errors. These disturbances are estimated in real time and actively compensated for in the control law. This approach enables the controller to remain effective even when precise modeling is difficult or when disturbances vary during operation.

While these methods are broadly applicable across robotic domains, the talk will focus on their application to Unmanned Aerial Vehicles (UAVs), which are particularly sensitive to disturbances like wind gusts and GPS noise. We will demonstrate how combining ILC with ESO-based disturbance rejection enhances trajectory tracking in UAVs during both repetitive missions and unpredictable environmental conditions. Specific examples include using ILC to improve flight accuracy in repetitive circular paths, and ESOs to mitigate errors introduced by delayed, degraded or falsified GPS signals.

Simulation results developed using MATLAB and Simulink will be presented, showing performance improvements compared to conventional PID-based control alone. The examples highlight both individual and combined effects of ILC and ESO, offering a practical roadmap for implementing robust control in robotic systems.

The techniques discussed in this session are not only relevant to UAVs but also generalizable to a wide range of robotics applications and platforms. Attendees will gain insights into how learning-based and observer-based control frameworks can work in tandem to address modern robotics challenges in real-time and uncertain environments.

Presenting Author: Siddharth Jawahar MathWorks Inc.

Presenting Author Biography: Siddharth Jawahar is a Customer Success Engineer at MathWorks, specializing in Model-Based Design. He focuses on the design and development of control algorithms for a wide range of applications in the automotive, aerospace, and defense industries. Siddharth holds a degree in Electrical Engineering with a specialization in control system design from the Georgia Institute of Technology.

Authors:

Siddharth Jawahar MathWorks Inc.