Identify a process to build a mathematical model from inputs/outputs recording
Specifying tests
Choose test scenarios – Choose sample time – Anti-aliasing filters – Conduct tests – Closed loop tests
Understanding numerical identification
Define identification goals – Analyze test results – Set numerical identification process – Closed-loop identification – Validate an identified model
Discovering identification methods
Simple least squares methods – Recursive least squares method – Unbiased estimators method – Instrumental variable methods – Generalised least squares method – Extended least squares method – Output error method
Identify nonlinear systems
Principle – IT structure to build – Application with an example