One common situation where you might want to include some discrete logic in an otherwise continuous model would be population dynamics. Using cohorts is a standard approach to doing this but something like (convey4a.mdl)
If the above model is formulated with the obvious equations such as
Maturing = young/first 16 years
Then, because of the nature of exponential decay, an increase in births at time 0 will lead to more young adults a year later. To overcome this we can reformulate with conveyor delays as (convey4.mdl).
This is almost the same except that maturing and aging are being computed using DELAY CONVEYOR. The behavior, with the same set of parameters, is significantly different:
