Ali BaniAsad

Robotics Engineer at Fasta | Robust RL & Embedded AI Researcher | Seeking PhD Positions

Publications | Ali BaniAsad

Publications


Ellipsoidal Set-Theoretic Design of Robust Safety Filters for Constrained Linear Systems

Reza Pordal, Alireza Sharifi, Ali Baniasad

arXiv preprint, 2025.

An ellipsoidal set-theoretic framework for designing robust safety filters that enforce state and input constraints on linear systems subject to bounded disturbances, providing formal safety guarantees for constrained control.

Applied an in-motion transfer alignment approach during global positioning system outages utilizing a recurrent neural network algorithm

Alireza Sharifi, Ali Baniasad, Saeid Mozafari

Engineering Applications of Artificial Intelligence, vol. 157, 2025.

A lightweight two-layer LSTM is trained online while GNSS is available, ingesting inertial measurements and wheel-odometry states. When the satellite link is lost, the network propagates position, velocity, and heading, constraining drift to less than 0.3% of distance travelled. In simulated 15-minute GNSS blackouts, lateral error remained below 25 cm...

Attitude control of a 3-DoF quadrotor platform using a linear quadratic integral differential game approach

Ali BaniAsad, Reza Pordal, Alireza Sharifi, Hadi Nobahari

ISA Transactions, vol. 148, pp. 515-527, 2024.

A quadcopter stand with three degrees of freedom is controlled using a differential-game framework. Player 1 tracks reference commands while Player 2 injects worst-case disturbances. The Nash-equilibrium solution yields a robust Linear Quadratic Integral Differential Game (LQIDG) controller with strong disturbance rejection. Performance is verified in MATLAB Simulink and on...

Linear Quadratic Integral Differential Game applied to the Real-time Control of a Quadrotor Experimental setup

Hadi Nobahari, Ali Baniasad, Alireza Sharifi

2022 10th RSI International Conference on Robotics and Mechatronics (ICRoM), 2022.

A linear quadratic differential game with integral action (LQIDG) is implemented in real time on a 3-DoF quadrotor stand. The two-player Nash-equilibrium formulation provides robust tracking of Euler-angle commands under worst-case disturbances, validated on physical hardware.