Deep neural network aided Monte Carlo simulation in solder joint failure probability analysis

In this paper, a novel method based on deep neural networks is proposed to consider the uncertainties of the influential parameters using Monte Carlo simulation. The method reduces the number of sampling by concen-trating the sampling range into the state limit function of the power converter of interest. The proposed method is applied to a PV inverter and reduces the calculation time to 25% of the so-called method.

Automated summary

This paper proposes a novel method based on deep neural networks to consider the uncertainties of influential parameters using Monte Carlo simulation. The method reduces the number of sampling by concentrating the sampling range into the state limit function of the power converter of interest. The proposed method is applied to a PV inverter and reduces the calculation time to 25% of the so-called method.