We study the time-dependent behavior of dust and plasma parameters for artificial electron heating experiments during Polar Mesospheric Summer Echoes (PMSE) conditions, by the use of the model for dusty plasmas introduced by Havnes and Morfill <1984> and Havnes et al. <1990, 1992>. We find that if the heater is run in a long series of cycles with equal and comparatively short (e.g., 10--20 s) intervals in the on and off mode, the difference between the dusty plasma parameters in subsequent heater off phases will be small. This will give the impression that the PMSE, which can be severely weakened during the heater on phase, returns more or less exactly to its value in the former heater off phase. However, we also find that if the heater off phase is kept long (e.g., 160 s), during which the dusty plasma in the radar beam will relax to its undisturbed value or be replaced by unaffected dusty plasma due to horizontal transport by wind, we will see a substantial difference in the plasma parameters from just before the heater is switched on and just after it is switched off. Owing to an additional charging of the dust during the heater on phase and which takes a long time to decharge, the negative charge state of the dust is higher when the heater is switched off than before it was on. This will increase electron gradients in the dusty plasma and also the value of dust charge density to electron density, and we propose that this should lead to a stronger PMSE, or an PMSE overshoot, when the heater is switched off, compared with the value just before it was switched on.