Performance evaluation of a boost converter under influence of random noise

The presence of unavoidable disturbances has become an issue concerning power converter modeling. These disturbances result in system parameter variations, which may significantly alter predicted performance of a converter. In this paper, a switching pulse-width modulated (PWM) DC-DC converter performance is studied under the influence of random noise. We experimentally verify existence of noise and measure its statistical properties on the operating characteristics of the boost converter. Experimental results show that the characteristics of random noise in the converter are Gaussian distributed. Available deterministic approaches do not provide insight into important effects of random noise on dynamics of a power converter. To reflect the system parameter uncertainties, these disturbances must be included in the converter model. To address this problem, we present a converter model that captures the effect of these inherent random noises on the overall behavior of a practical converter. Realistic disturbances are lumped and modeled as switching time uncertainties. We also propose a performance index quantifying the converter performance. This index is known as mean first passage time (MFPT). It represents the time for the trajectory to evolve from operating point to the stability boundary. Comparison between stochastic performance index, MFPT and the deterministic one, critical energy, under the presence of realistic disturbances is presented. The MFPT illustrates the use of proposed stochastic model to determine the power converter limit.