Determining the swarm parameters of gases considering ion kinetics by parallel genetic algorithm on GPU platform

In this work, a convenient and efficient method is proposed to determine swarm parameters considering ion kinetics from pulsed Townsend (PT) measurements. First, a physical model was presented to describe the development of PT discharge considering electron detachment and ion conversion reactions. A numerical solution to the model was also proposed. In order to assess the precision of our calculations, we presented the calculated electronic and ionic transients derived from our model for different cases. Then, a genetic algorithm (GA) was proposed to find a set of swarm parameters, under which the deviation between the simulated current waveform and the actual measured current waveform is minimum. It is time-consuming to simulate a single waveform, and since a large number of waveforms need to be simulated in the GA, graphic processing unit-based parallel computing is used to improve computing efficiency. Finally, the swarm parameters of dry air considering electron detachment and ion conversion processes using the method were obtained and they are in good agreement with those in references.