Optimisation and System Identification
University of Strathclyde, Glasgow, UK,
7th - 9th May 2008
This meeting has since been held. (Register your interest for the next course here.)
This three-day course introduces two closely related topics of Optimisation and System Identification.
The Optimisation is extremely important subject used across all the Engineering fields. For control engineering applications it is widely used for optimal control and parameter identification/estimation. The system identification is probably the most important and difficult step required for a successful modern control design.
The course is aimed at engineers who are involved in system modeling and model based control/simulation.
All presented topics will be supported by practical engineering examples and will arm trainees with efficient approach to tackle real-life problems. Lectures and Hands-On sessions will provide a methodology and step-by-step guide of using all presented algorithms in their engineering practice.
|Day 1: Optimisation|
|09.00||Static optimisation methods: Unconstrained Optimisation|
|09.45||Static optimisation methods: Constrained Optimisation|
|11.00||Hands-on: Static Optimisation Methods|
|13.30||Dynamic Optimisation Methods|
|14.30||Hands-On: Dynamic Optimisation Methods|
|15.50||Genetic and Search Algorithms|
|16.30||Hands-On: Genetic/Search algorithms|
|Day 2: Linear Systems Identification|
|09.00||System Identification for Linear Systems|
|11.00||Hands-on: System ID with Least Squares Algorithm|
|11.45||Hands-on: System ID with a Kalman Filter to Estimate an Offset|
|13.15||System Identification Implementation Issues and Model Validation|
|14.30||Hands-On: Estimation of Parameters in Physics-based Models|
|Day 3: Non-Linear System Identification|
|09.00||Grey-Box models - Physical System Structure with Black-Box Elements||10.00||Parameter Estimation for Grey-Box Models|
|11.15||Hands-on: Grey-Box Model Identification|
|13.15||Non-linear System Modeling through Multiple Linear Models|
|14.15||Hands-On: Multiple-Model Approach|
|15.30||Neural Networks as Universal Approximators: Static and Dynamic Models|
|16.30||Demonstration of the Neural-Network Model Identification Procedure|