Mohit Mehndiratta
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Mohit Mehndiratta has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
- Erdal Kayacan
Automatic Control Group (RAT), Department of Electrical Engineering and Information Technology, Paderborn University, Paderborn, Germany
4 shared publications - Efe Camci
School of Mechanical and Aerospace Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Singapore
1 shared publication - Girish Chowdhary
University of Illinois at Urbana-Champaign (UIUC)
1 shared publication - Mahmut Reyhanoglu
College of Science and Engineering, University of Houston - Clear Lake, Houston, United States, USA
1 shared publication Siddharth Patel
Nanyang Technological University
1 shared publication
Latest Publications
- Open Access
- 7 September 2021
- Published by MDPI AG in Electronics
- Vol. 10 (18) , 2187
Can Deep Models Help a Robot to Tune Its Controller? A Step Closer to Self-Tuning Model Predictive Controllers
Motivated by the difficulty roboticists experience while tuning model predictive controllers (MPCs), we present an automated weight set tuning framework in this work. The enticing feature of the proposed methodology is the active exploration approach that adopts the exploration–exploitation concept at its core. Essentially, it extends the trial-and-error method by benefiting from the retrospective knowledge gained in previous trials, thereby resulting in a faster tuning procedure. Moreover, the tuning framework adopts a deep neural network (DNN)-based robot model to conduct the trials during the simulation tuning phase. Thanks to its high fidelity dynamics representation, a seamless sim-to-real transition is demonstrated. We compare the proposed approach with the customary manual tuning procedure through a user study wherein the users inadvertently apply various tuning methodologies based on their progressive experience with the robot. The results manifest that the proposed methodology provides a safe and time-saving framework over the manual tuning of MPC by resulting in flight-worthy weights in less than half the time. Moreover, this is the first work that presents a complete tuning framework extending from robot modeling to directly obtaining the flight-worthy weight sets to the best of the authors’ knowledge.
- 7 September 2021
- Published by MDPI AG in Electronics
- Vol. 10 (18) , 2187
Open AccessCan Deep Models Help a Robot to Tune Its Controller? A Step Closer to Self-Tuning Model Predictive Controllers
Motivated by the difficulty roboticists experience while tuning model predictive controllers (MPCs), we present an automated weight set tuning framework in this work. The enticing feature of the proposed methodology is the active exploration approach that adopts the exploration–exploitation concept at its core. Essentially, it extends the trial-and-error method by benefiting from the retrospective knowledge gained in previous trials, thereby resulting in a faster tuning procedure. Moreover, the tuning framework adopts a deep neural network (DNN)-based robot model to conduct the trials during the simulation tuning phase. Thanks to its high fidelity dynamics representation, a seamless sim-to-real transition is demonstrated. We compare the proposed approach with the customary manual tuning procedure through a user study wherein the users inadvertently apply various tuning methodologies based on their progressive experience with the robot. The results manifest that the proposed methodology provides a safe and time-saving framework over the manual tuning of MPC by resulting in flight-worthy weights in less than half the time. Moreover, this is the first work that presents a complete tuning framework extending from robot modeling to directly obtaining the flight-worthy weight sets to the best of the authors’ knowledge.
- Open Access
- 27 December 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 1653-1669
Robust Tracking Control of Aerial Robots Via a Simple Learning Strategy-Based Feedback Linearization
To facilitate accurate tracking in unknown/uncertain environments, this paper proposes a simple learning (SL) strategy for feedback linearization control (FLC) of aerial robots subject to uncertainties. The SL strategy minimizes a cost function defined based on the closed-loop error dynamics of the nominal system via the gradient descent technique to find the adaptation rules for feedback controller gains and disturbance estimate in the feedback control law. In addition to the derivation of the SL adaptation rules, the closed-loop stability for a second-order uncertain nonlinear system is proven in this paper. Moreover, it is shown that the SL strategy can find the global optimum point, while the controller gains and disturbance estimate converge to a finite value which implies a bounded control action in the steadystate. Furthermore, utilizing a simulation study, it is shown that the simple learning-based FLC (SL-FLC) framework can ensure desired closed-loop error dynamics in the presence of disturbances and modeling uncertainties. Finally, to validate the SL-FLC framework in real-time, the trajectory tracking problem of a tilt-rotor tricopter unmanned aerial vehicle under uncertain conditions is studied via three case scenarios, wherein the disturbances in the form of mass variation, ground effect, and wind gust, are induced. The real-time results illustrate that the SL-FLC framework facilitates a better tracking performance than that of the traditional FLC method while maintaining the nominal control performance in the absence of modeling uncertainties and external disturbances, and exhibiting robust control performance in the presence of modeling uncertainties and external disturbances.
- 27 December 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 1653-1669
Open AccessRobust Tracking Control of Aerial Robots Via a Simple Learning Strategy-Based Feedback Linearization
To facilitate accurate tracking in unknown/uncertain environments, this paper proposes a simple learning (SL) strategy for feedback linearization control (FLC) of aerial robots subject to uncertainties. The SL strategy minimizes a cost function defined based on the closed-loop error dynamics of the nominal system via the gradient descent technique to find the adaptation rules for feedback controller gains and disturbance estimate in the feedback control law. In addition to the derivation of the SL adaptation rules, the closed-loop stability for a second-order uncertain nonlinear system is proven in this paper. Moreover, it is shown that the SL strategy can find the global optimum point, while the controller gains and disturbance estimate converge to a finite value which implies a bounded control action in the steadystate. Furthermore, utilizing a simulation study, it is shown that the simple learning-based FLC (SL-FLC) framework can ensure desired closed-loop error dynamics in the presence of disturbances and modeling uncertainties. Finally, to validate the SL-FLC framework in real-time, the trajectory tracking problem of a tilt-rotor tricopter unmanned aerial vehicle under uncertain conditions is studied via three case scenarios, wherein the disturbances in the form of mass variation, ground effect, and wind gust, are induced. The real-time results illustrate that the SL-FLC framework facilitates a better tracking performance than that of the traditional FLC method while maintaining the nominal control performance in the absence of modeling uncertainties and external disturbances, and exhibiting robust control performance in the presence of modeling uncertainties and external disturbances.
- 2 July 2019
- Published by Springer Nature in Autonomous Robots
- Vol. 43 (8) , 2209-2228
A constrained instantaneous learning approach for aerial package delivery robots: onboard implementation and experimental results
- 2 July 2019
- Published by Springer Nature in Autonomous Robots
- Vol. 43 (8) , 2209-2228
A constrained instantaneous learning approach for aerial package delivery robots: onboard implementation and experimental results
- 1 September 2018
- Published by Springer Nature
- p. 581-605
Learning-Based Fast Nonlinear Model Predictive Control for Custom-Made 3D Printed Ground and Aerial Robots
- 1 September 2018
- Published by Springer Nature
- p. 581-605
Learning-Based Fast Nonlinear Model Predictive Control for Custom-Made 3D Printed Ground and Aerial Robots