IPSO-based Parameter Identification of Semiconductor Devices

An efficient and simple method of parameter identification for semiconductor devices is presented in this paper. The method is based on improved particle swarm optimization (IPSO), which can overcome some deficiencies of standard PSO. A typical semiconductor device, Schottky-barrier diode (SBD), has been used as an example for demonstration. The performance of the IPSO was compared with the standard PSO and the genetic algorithm (GA) that is commonly applied to solve parameter identification problems of semiconductor devices. The comparative results indicate that the IPSO method can obtain optimum solutions more easily than others.