How to forecast N-steps ahead in System Identification Toolbox for Hammerstein Wiener model

In the System identification Toolbox I can create a Hammerstein-Wiener model for certain data. After the estimation I can use the model in the “prediction” function but not in the “forecast” function. Then I get an error message.

The goal that I want to forecast N-steps ahead, so outside the input data range.

How can this be solve? What is the correct way to come to N-steps forecast for y(t+N)?