Developed a set of adaptive MPC algorithms for online model learning during MPC design. The algorithms are valid for both parametric and non-parametric model uncertainties.
Monimoy Bujarbaruah, Xiaojing Zhang, Marko Tanaskovic, Francesco Borrelli; Adaptive Stochastic MPC under Time Varying Uncertainty, IEEE Transactions on Automatic Control, June 2021 [pdf]
Monimoy Bujarbaruah, Siddharth H. Nair, Francesco Borrelli; A Semi-Definite Programming Approach to Robust Adaptive MPC under State Dependent Uncertainty; IEEE European Control Conference (ECC), May 2020 [pdf, code]
Siddharth H. Nair, Monimoy Bujarbaruah, Francesco Borrelli; Modeling of Dynamical Systems via Successive Graph Approximations; IFAC World Congress, July 2020 [pdf]
Monimoy Bujarbaruah, Charlott Vallon; Exploiting Model Sparsity in Adaptive MPC: A Compressed Sensing Viewpoint, Learning for Dynamics and Control (L4DC), June 2020 [pdf]
Monimoy Bujarbaruah, Xiaojing Zhang, Ugo Rosolia, Francesco Borrelli; Adaptive MPC for Iterative Tasks, IEEE Conference on Decision and Control (CDC), Miami, FL, December 2018 [pdf]
Monimoy Bujarbaruah, Xiaojing Zhang, H. Eric Tseng, Francesco Borrelli; Adaptive MPC for Autonomous Lane Keeping, 14th International Symposium on Advanced Vehicle Control (AVEC), Beijing, China, July 2018 [pdf]
Monimoy Bujarbaruah, Xiaojing Zhang, Francesco Borrelli; Adaptive MPC with Chance Constraints for FIR Systems, IEEE American Control Conference (ACC), Milwaukee, WI, June 2018 [pdf]
Developed a set of algorithms for learning and adapting state-input constraints during MPC design.
The input constraint adapting algorithms were subsequently utilized in a trajectory optimization problem and experimentally tested on a full-scale autonomous truck.
Relevant Papers
Monimoy Bujarbaruah, Charlott Vallon, Francesco Borrelli; Learning to Satisfy Unknown Constraints in Iterative MPC, IEEE Conference on Decision and Control (CDC), Jeju, South Korea, December 2020 [pdf, code]
Lars Svensson, Monimoy Bujarbaruah, Nitin Kapania, Martin Törngren; Adaptive Trajectory Planning and Optimization at Limits of Handling; IEEE International Conference on Intelligent Robots and Systems (IROS), Macau, November 2019 [pdf]
Lars Svensson, Monimoy Bujarbaruah, Arpit Karsolia, Christian Berger, Martin Törngren; Traction Adaptive Motion Planning at the Limits of handling, IEEE Transactions on Control System Technology [pdf]
Developed two safe and near optimal primal-dual MPC policy learning algorithms for MPC using deep neural networks.
The developed algorithms were implemented on vehicle chassis control and lane keeping applications using ROS to obtain 10x speedup over the Gurobi solver.
Relevant Papers
Xiaojing Zhang*, Monimoy Bujarbaruah*, Francesco Borrelli; Near-Optimal Rapid MPC using Neural Networks: A Primal-Dual Policy Learning Framework, IEEE Transactions on Control System Technology (*equal contribution) [pdf]
Xiaojing Zhang, Monimoy Bujarbaruah, Francesco Borrelli; Safe and Near-Optimal Policy Learning for Model Predictive Control using Primal-Dual Neural Networks; IEEE American Control Conference (ACC), Philadelphia, PA, July 2019 [pdf, code]
Developed an algorithm for online learning of disturbance distribution bounds using Bootstrap confidence intervals.
Deployed the algorithm with an MPC controller on a UR5e robot which learned to solve the ball-in-cup problem.
Relevant Papers
Monimoy Bujarbaruah, Akhil Shetty, Kameshwar Poolla, Francesco Borrelli; Learning Robustness with Bounded Failure: An Iterative MPC Approach, IFAC World Congress, July 2020 [pdf, code]
Monimoy Bujarbaruah, Tony Zheng, Akhil Shetty, Martin Sehr, Francesco Borrelli; Learning to Play Cup-and-Ball with Noisy Camera Observations, IEEE International Conference on Automation Science and Engineering (CASE), Hong Kong, August 2020 [pdf, code]
Developed a set of robust MPC algorithms that handle both parametric model uncertainties and additive disturbances in the system.
The algorithms obtained an improved computational complexity vs conservatism trade-off over state-of-the-art tube MPC approaches across various numerical examples. See this repository for examples.
Relevant Papers
Monimoy Bujarbaruah, Ugo Rosolia, Yvonne R. Stürz, Xiaojing Zhang, Francesco Borrelli; Robust MPC for LPV Systems via a Novel Optimization-Based Constraint Tightening, Automatica, September 2022 [pdf, code]
Monimoy Bujarbaruah, Ugo Rosolia, Yvonne R. Stürz, Francesco Borrelli; A Simple Robust MPC for Linear Systems with Parametric and Additive Uncertainty, IEEE American Control Conference (ACC), May 2021 [pdf, code]
Developed an algorithm for learning unknown obstacles and safely completing a transportation task with two robots.
The control algorithm is decentralized, and the robots switch leader-follower roles with only local estimates, and implicit communications such as force and torque feedback.
Relevant Papers
Solving the physics based puzzle Bubble Ball, a convex optimization based planner-tracker synthesis for nonlinear systems, lane change assistance controller for vehicles with active front steering, and an attitude constrained controller for a spacecraft.
Relevant Papers
He Yin*, Monimoy Bujarbaruah*, Murat Arcak, Andrew Packard; Optimization Based Planner-Tracker Design for Safety Guarantees; IEEE American Control Conference (ACC), July 2020 (*equal contribution) [pdf]
Hotae Lee, Monimoy Bujarbaruah, Francesco Borrelli; Learning How to Solve Bubble Ball, Learning for Dynamics and Control (L4DC), June 2021 [pdf, video]
Monimoy Bujarbaruah, Ziya Ercan, Vladimir Ivanovic, H. Eric Tseng, Francesco Borrelli; Torque Based Lane Change Assistance with Active Front Steering, IEEE International Conference on Intelligent Transportation Systems (ITSC), Yokohama, Japan, October 2017 [pdf]
Monimoy Bujarbaruah, Srikant Sukumar; Lyapunov Based Attitude Constrained Control of a Spacecraft, AAS-AIAA Astrodynamics Specialist Conference, Vail-CO, USA, August 2015 [pdf]
