Behavioural Theory & Data-Driven Control Course | Prof. Debasattam Pal IIT Bombay
Course Details
| Exam Registration | 26 |
|---|---|
| Course Status | Ongoing |
| Course Type | Elective |
| Language | English |
| Duration | 8 weeks |
| Categories | Electrical, Electronics and Communications Engineering, Control and Instrumentation |
| Credit Points | 2 |
| Level | Postgraduate |
| Start Date | 16 Feb 2026 |
| End Date | 10 Apr 2026 |
| Enrollment Ends | 16 Feb 2026 |
| Exam Registration Ends | 27 Feb 2026 |
| Exam Date | 18 Apr 2026 IST |
| NCrF Level | 4.5 — 8.0 |
Behavioural Theory of Systems with a View Toward Data Driven Control
The landscape of control systems engineering is undergoing a paradigm shift. The traditional model-based approach, reliant on first-principles or system identification, is being complemented by powerful data-driven methodologies. At the forefront of this revolution lies the behavioural approach to systems theory, a framework uniquely suited for harnessing data directly for control design. This 8-week postgraduate course, instructed by Prof. Debasattam Pal of IIT Bombay, offers a deep dive into this powerful synergy.
About the Course
Data-driven control is an emerging paradigm that enables the design of robust control systems without explicitly constructing a mathematical model. Instead of focusing on model parameters, it leverages the system's observed "behaviour"—the set of all possible trajectories it can produce. This course bridges the foundational concepts of behavioural systems theory with cutting-edge data-driven control techniques. You will start by mastering the behavioural framework, which places system trajectories at the centre-stage, and then apply these principles to understand and implement data-driven control strategies, anchored by the fundamental concept of persistency of excitation.
Course Instructor: Prof. Debasattam Pal
Prof. Debasattam Pal is an Associate Professor in the Department of Electrical Engineering at IIT Bombay. An alumnus of IIT Bombay (MTech 2007, PhD 2012), he has established himself as a leading researcher with over 70 publications in top-tier journals and conferences. His expertise spans algebraic-geometric analysis of systems, dissipativity theory, switched systems, optimal control, and multi-agent systems, providing a rich and authoritative foundation for this course.
Who Should Enroll?
Intended Audience: This course is designed for postgraduates and professionals with a background in control theory. Ideal candidates include:
- Engineers with a BE/BTech degree who have completed an undergraduate course in control theory.
- MSc graduates in Mathematics with coursework in dynamical systems.
Prerequisites: A solid understanding of basic linear algebra and classical control theory is required to fully engage with the course material.
Industry Relevance
The skills acquired in this course are highly valued across industries that rely on advanced control systems design. Companies in aerospace, robotics, automotive, chemical processing, and industrial automation actively seek expertise in modern, data-centric control methodologies. This course is recognized as relevant for professionals at leading firms such as Eaton, Honeywell, GE, and ABB.
Detailed Course Layout
| Week | Topics Covered |
|---|---|
| Week 1 | Introduction: Systems, behaviours, dynamical systems. Fundamental properties: linearity, shift/time-invariance. |
| Week 2 | Models and representations: AR, MA, ARMA. Kernel, image, latent variables. |
| Week 3 | Behaviours and modules: The fundamental principle and elimination theory. |
| Week 4 | Two crucial concepts: Controllability and observability. |
| Week 5 | From time series to linear systems: the most powerful unfalsified model. |
| Week 6 | The fundamental lemma of data-driven approach: persistency of excitation. |
| Week 7 | Data-driven control: Data-driven stabilization. |
| Week 8 | Data-driven control (contd.): Data informativity. |
Key Learning Outcomes
- Grasp the core principles of the behavioural approach to systems and control.
- Understand the mathematical representation of system behaviours using AR/MA/ARMA models.
- Analyze fundamental system properties like controllability and observability from a behavioural perspective.
- Learn to construct models directly from time-series data.
- Master the critical concept of persistency of excitation and its role in data-driven control.
- Apply behavioural theory to design data-driven controllers for system stabilization.
- Evaluate the quality and sufficiency of data for control objectives using data informativity theory.
Recommended Textbooks & References
- Polderman and Willems: Systems Theory: a Behavioral Approach – The foundational textbook for behavioural theory.
- Willems et al.: “A note on persistency of excitation” – A seminal paper introducing the key lemma for data-driven methods.
- de Persis and Tesi: “Formulas for Data-Driven Control: Stabilization, Optimality, and Robustness” – A comprehensive guide to modern data-driven control formulas.
- van Waarde et al.: “Data Informativity: A New Perspective on Data-Driven Analysis and Control” – Explores the theory behind assessing data quality for control.
This course represents a unique opportunity to transition from classical control paradigms to the data-driven future, guided by one of India's premier institutions and an expert in the field. Enroll to build a sophisticated understanding of how to let data guide the control of complex systems.
Enroll Now →