The data is sampled at an interval given by the software system.
According to the Nyquist sampling theorem, Jerri (1977), if, to retain the information that a continuous signal contains, the sampling rate must be greater than twice the highest frequency component in the original signal to avoid frequency aliasing ( = confusion between low- and high-frequency components in the original data), the Nyquist frequency is half of the sampling rate of the discrete time series.
Frequency aliasing can be removed by introducing a low-pass bandwidth filter.
At least two sample values per cycle are required to define the highest frequency in the signal.
If the sampling rate is larger than the pulse period of the signal there is a risk of losing information about the frequency of the signal.
When determining the sample rate it is therefore necessary to have information about expected frequencies in the measured data, e.g. expected bow motion measured by an accelerometer or expected cycles from the ship’s rudder.
In order to identify the resistance and to implement a rudder resistance model, it is considered sufficient with a 10 seconds update on the rudder movements.
The suggested logging period of all parameters is per default set to 10 seconds.