Linear Systems. Analysis of Complex Nonlinear Data Sets and Model Development. Meta Modelling vs Mechanistic Modelling
Abstract:
The aim of the project is to introduce students in to pragmatic engineering mind set. The first part will introduce students to process of model construction of simple mechanical/electrical system for the purpose of building an understanding how input output relationship is derived using first principles of mechanics (Newtonian Mechanics). First two models will be of a mechanical pendulum and a mass spring damper system (MSD). The later part will demonstrate the process of modelling a complex mechanical system such as robot manipulator. The next part will be focusing on lack of knowledge that engineers and scientists face during model construction and how data driven modelling approach combined with initial mathematical formulation of a system under consideration can help improve initial model and improve control of complex systems. Mechanical systems even though complex, they are orders of magnitude simpler than biological systems with multiple inputs and multiple outputs and number of hidden states. Thus, having gained knowledge in to control engineering perspective of analysis of a mechanical system from inputoutputperspective, students should be able to formulate pragmatic approach when dealing with complex multiple input multiple output biological systems such as human brain. Finally, the last part will introduce students to brain theory as a system of meta-model construction and world interpretation
Project prerequisites:
- Matrix Algebra / Linear Algebra
- Physics (High School)
- MATLAB / C++ / Whatever Programming language
Planned lectures:
- Dynamic Modelling and Control of Mechanical Systems (Pendulum, MSD, 6DOF ABB IRB140 Manipulator)
- Multivariate Data Analysis (SVD, PCA, ICA, PLSR)
- Model Development with MVA and Newtonian Mechanics
Associated topics:
Linear Systems, Nonlinear Systems, Multivariate Data Analysis, Systems Engineering, Modelling and Control of Dynamic Systems, Feedback Control, Adaptive Control, Neural Networks.
About lecturer:
Mr. Serge Gale,
PhD Candidate in Engineering Cybernetics and Machine Intelligence at Institute of Technical Cybernetics (ITK), NTNU, Trondheim