AI-based Nonlinear Model Predictive Control for Flying Robots

Last date of Registration : 3rd December, 2025

SKU: Date : 6th December, 2025 Category: Short Term Programs: CLOSED

Description about the workshop

This workshop provides comprehensive, hands-on experience designing a nonlinear model predictive controller (NLMPC) with a neural state-space prediction model for a free-flying robot. Throughout the workshop, participants use MATLAB to implement the control strategy. The first step is to identify the neural state-space model by loading a dataset of state and input trajectories and training a multi-layer perceptron (MLP) network on this data. Once the neural state-space model is identified, participants estimate the unmeasured states of the flying robot model using an extended Kalman filter. Next, participants plan optimal state and input trajectories using a multistage NLMPC that employs the previously identified neural state-space model as its predictor. Finally, participants simulate the NLMPC in closed loop with the flying robot model and plot the robot’s actual state trajectories and inputs against the optimal ones to confirm successful tracking.

Profile of the instructors

Ms. Sragdhara is an online content developer at MathWorks focused on developing Simulink-based learning content. She holds a PhD degree in Control Systems. Her research areas include robust control, multirate control, cooperative control of multi-agent systems, decentralized control, and control of cyber-physical systems. She has published papers in international journals and conferences in the domain of her research. She is also an AI enthusiast and is currently exploring the applications of machine learning and deep learning in the areas of signal and image processing, and developing course content.

Ms. Shanthi is an experienced educator and online content developer at MathWorks, specializing in MATLAB for advanced signal processing and AI applications. With a master’s degree in Signal Processing, she brings deep technical expertise in Digital Signal Processing (DSP), Image Processing, Adaptive Filtering, Wireless Communication Systems, and the application of Neural Networks to signal and image data. Her scholarly contributions include publications in international journals on topics such as Video Compression Standards and their implementation in MATLAB, and Background Subtraction Techniques in Image Processing. She is skilled in algorithm development for signal enhancement, noise reduction, feature engineering, and real-time data analysis, with a strong command of MATLAB toolboxes for signal, image, and communication processing. At MathWorks, she creates engaging, practical online courses that help learners apply MATLAB to solve real-world signal processing and AI challenges.

Modules of the workshop

Module name	Concepts covered	Live sessions - No. of hours	Learning outcomes
Getting Started with MATLAB Online	NA	Live Session 10 minutes	NA
Overview of Nonlinear Model Predictive Control (NLMPC) with neural state-space prediction model	What is MPC and NLMPC? What is neural state-space? What is NLMPC with neural state-space?	Live Session 20 minutes	Describe the use of neural state-space models in Nonlinear Model Predictive Control (NLMPC).
Identify a neural state-space prediction model	Training a neural state-space	Live Session 15 minutes	Construct a neural state-space prediction model using a multi-layer perceptron (MLP) network. Train the MLP network with specified training options for state-space prediction.
Design NLMPC with neural state-space prediction model	How to use an identified neural state-space model as prediction model in NLMPC	Live Session 35 minutes	Implement a nonlinear model predictive controller (NLMPC) using a neural state-space prediction model and its analytical Jacobian. Configure key NLMPC properties, including constraints, horizons, sample time, cost function, and input bounds.
Nonlinear estimation of unmeasurable states using extended Kalman filter (EKF)	Estimating unmeasured states using an EKF	Live Session 25 minutes	Implement an EKF by configuring the filter object and setting measurement noise parameters.
Optimal trajectory planning	Planning an optimal trajectory (for the robot to track) using a multistage NLMPC with the identified neural state-space prediction model	Live Session 30 minutes	Implement a multistage NLMPC using a neural state-space prediction model and visualize the resulting optimal state trajectories and control inputs.
Simulate the NLMPC with a neural state-space prediction model	Run closed-loop simulation of an NLMPC with a neural state-space prediction model	Live Session 20 minutes	Simulate the NLMPC by updating state estimates using the EKF based on plant measurements. Evaluate tracking performance by comparing and visualizing actual and optimal state trajectories and control inputs.
Summarize	NA	Live Session 5 minutes	NA
Q & A	NA	Live Session 10 minutes	NA

Session Details

Date of the Workshop : 6th December, 2025

Timings (IST) : 10:00 a.m. – 01:00 p.m.

Mode of the Workshop : Online

Intended Audience, Eligibility, Prerequisites & Tools/software used

Intended Audience :

Engineering students, Researchers and Industry Professionals working on Robotics, control systems and AI - based control.

Eligibility :

UG or PG Engineering students, Researchers, and Industry Professionals working on robotics, control systems and AI-based control.

Prerequisites :

Basic MATLAB and Simulink knowledge.

Tools/software that will be used :

MATLAB (Participants will be given a free License to the product for 1 Week.)

Fees for the workshop

Students : Rs. 236/- ( Rs. 200 + 18% GST)

Faculty : Rs. 590/- ( Rs. 500 + 18% GST)

Industry Professionals : Rs. 1180/- ( Rs. 1000 + 18% GST)

Certification Criteria

Attending the workshop is mandatory for certification.

Reviews

There are no reviews yet.

Be the first to review “AI-based Nonlinear Model Predictive Control for Flying Robots”

Cart