arXiv subjects

Deepak Ingole

Publications and source records attributed to Deepak Ingole.

4 recordsLinked to original sources

arXiv2025

Iterative Tuning of Nonlinear Model Predictive Control for Robotic Manufacturing Tasks

Manufacturing processes are often perturbed by drifts in the environment and wear in the system, requiring control re-tuning even in the presence of repetitive operations. This paper presents an iterative learning framework for automatic tuning of Nonlinear Model Predictive Control (NMPC) weighting matrices based on task-level performance feedback. Inspired by norm-optimal Iterative Learning Control (ILC), the proposed method adaptively adjusts NMPC weights Q and R across task repetitions to minimize key performance indicators (KPIs) related to tracking accuracy, control effort, and saturation. Unlike gradient-based approaches that require differentiating through the NMPC solver, we construct an empirical sensitivity matrix, enabling structured weight updates without analytic derivatives. The framework is validated through simulation on a UR10e robot performing carbon fiber winding on a tetrahedral core. Results demonstrate that the proposed approach converges to near-optimal tracking performance (RMSE within 0.3% of offline Bayesian Optimization (BO)) in just 4 online repetitions, compared to 100 offline evaluations required by BO algorithm. The method offers a practical solution for adaptive NMPC tuning in repetitive robotic tasks, combining the precision of carefully optimized controllers with the flexibility of online adaptation.

cs.RO↗

arXiv2025

Bayesian Optimization for Automatic Tuning of Torque-Level Nonlinear Model Predictive Control

This paper presents an auto-tuning framework for torque-based Nonlinear Model Predictive Control (nMPC), where the MPC serves as a real-time controller for optimal joint torque commands. The MPC parameters, including cost function weights and low-level controller gains, are optimized using high-dimensional Bayesian Optimization (BO) techniques, specifically Sparse Axis-Aligned Subspace (SAASBO) with a digital twin (DT) to achieve precise end-effector trajectory real-time tracking on an UR10e robot arm. The simulation model allows efficient exploration of the high-dimensional parameter space, and it ensures safe transfer to hardware. Our simulation results demonstrate significant improvements in tracking performance (+41.9%) and reduction in solve times (-2.5%) compared to manually-tuned parameters. Moreover, experimental validation on the real robot follows the trend (with a +25.8% improvement), emphasizing the importance of digital twin-enabled automated parameter optimization for robotic operations.

cs.RO↗

arXiv2022

Embedded Model Predictive Control Using Robust Penalty Method

Model predictive control (MPC) has become a hot cake technology for various applications due to its ability to handle multi-input multi-output systems with physical constraints. The optimization solvers require considerable time, limiting their embedded implementation for real-time control. To overcome the bottleneck of traditional quadratic programming (QP) solvers, this paper proposes a robust penalty method (RPM) to solve an optimization problem in a linear MPC. The main idea of RPM is to solve an unconstrained QP problem using Broyden Fletcher Goldfarb Shannon (BFGS) algorithm. The beauty of this method is that it can find optimal solutions even if initial conditions are in an infeasible region, which makes it robust. Moreover, the RPM is computationally inexpensive as compared to the traditional QP solvers. The proposed RPM is implemented on resource-limited embedded hardware (STM32 microcontroller), and its performance is validated with a case study of a citation aircraft control problem. We show the hardware-in-the-loop co-simulation results of the proposed RPM and compared them with the active set method (ASM) and interior point method (IPM) QP solvers. The performance of MPC with the aforementioned solvers is compared by considering the optimality, time complexity, and ease of hardware implementation. Presented results show that the proposed RPM gives the same optimality as ASM and IPM, and outperforms them in terms of speed.

eess.SY↗

arXiv2021

Embedded Model Predictive Controller on Low-Cost Low-End Microcontroller for Electrical Drives

It is very well-known that the implementation of Model Predictive Controller (MPC) on embedded platforms is challenging due to the computational complexities associated while solving an optimization problem. Although, there are many efficient embedded implementations existing by now, but for faster, more dynamic and non-linear control applications, there is no cost-effective and memory efficient embedded solutions. In this paper, we show the implementation of embedded explicit MPC for a motor speed control application on a lowcost 8 bit PIC 18 series microcontroller which costs only $5. The offset-free explicit MPC is designed for reference tracking, constraints handling, and disturbance rejection. The developed control law is exported to low-level C code and utilized in HIL co-simulations. We present the results of memory demand and control performance under various operating scenarios. The presented results show that the developed embedded MPC utilize about 40% of RAM and 92% of ROM for prediction horizon up to 3 samples. The performance of developed MPC is compared with the conventional PI controller. Overall results show that the presented approach is cost-effective, portable, and gives better performance than the PI controller.

eess.SY↗