| System Identification Toolbox | |
Compute the prediction errors associated with a model and a data set.
Syntax
Description
data is the output-input data set, given as an iddata object, and m is any idmodel object.
e is returned as an iddata object, so that e.OutputData contains the prediction errors that result when model m is applied to the data:
The argument init determines how to deal with the initial conditions:
init ='estimate' means that the initial state is chosen so that the norm of prediction error is minimized. This initial state is returned as x0.
init = 'zero' sets the initial state to zero.
init = 'model' used the model's internally stored initial state.
init = x0, where x0 is a column vector of appropriate dimension uses that value as initial state.
If init is not specified, the model property m.InitialState is used, so that 'Estimate', 'Backcast' and 'Auto' set init = 'Estimate', while m.InitialState = 'Zero' sets init = 'zero', and 'Fixed' and 'Model' set init = 'model'.
The output argument x0 is the used value of the initial state. If data contains several experiments, x0 will be a matrix, containing the initial states from each experiment.
See Also
| | oe | pem | |