Seminar 12(+13)

On 7.1 I say that we cannot derive the disturbance signal d. This is partly incorrect - we can derive it, but we would like to avoid doing so, since it makes the compensation very sensitive to measurement noise and quickly varying disturbances.