Integrate Eli Pulse train with randomised periods

[pausz]

This is an excellent idea for a stimulus and one that is often seen in experimental protocols. I guess thta's why @timysnoturtle is using it.

The method needs to be iterative. Thus, this issue probably links to #82 when we implement the new bandlimited noise method.

Pulse train properties that could be randomised:

• amplitudes
• periods (probably defined as the time between the start of a pulse and the start of the next pulse)
• duty cycle (percentage of time ON with respect to the period)

Square waveform

Rectangular waveform