Spread Spectrum in Random PWM DC-DC Converters by PSO&GA Optimized Randomness Levels

This paper proposes an optimized spread spectrum Random PWM technique for the reduction of Electromagnetic Interference (EMI) in DC-DC converter to improve the Electromagnetic Compatibility (EMC) of the system. In particular, the Randomized Pulse Position Modulation (RPPM) is considered and a study of the random laws is developed. Different density probability distributions are analyzed and the randomness levels, that characterize each probability distribution, are optimized in order to reduce the Power Spectral Density (PSD) of the DC-DC converter output voltage. Two optimization techniques are used: the Particle Swarm Optimization (PSO) technique and the Genetic Algorithm (GA). Four probability density distributions are analyzed, Uniform, Normal, Beta and Generalized Extreme Value Distribution (GEV) distributions, and their effectiveness on spreading the PSD is analyzed. Finally, the effects on the reduction of the output EMI of the converter is shown.