The droop control method is usually employed in the DC microgrids to share the load current demand among multiple sources due to its advantage of being independent of a communication network. However, the performance of the droop control method is affected by the mismatched transmission line resistance and the offset in the nominal voltage reference. This paper presents the design and optimization of the droop coefficient of converters, using the genetic algorithm to enhance the current sharing and the DC bus voltage regulation performance. The proposed approach is tested on the single bus multi-source electrical power system (EPS) for the more electric aircraft (MEA) applications. The effectiveness of the proposed approach is validated using a detailed simulation model of the MEA EPS developed in MATLAB Simulink.

Most of the existing methods used to estimate the cable resistance require the use of many hardware devices and the injection of perturbations to the system. Therefore, they are time-consuming, costly and prone to errors. In addition, the injection of perturbations has the potential of degrading the power quality of the system. In this paper, a new artificial neural network (ANN) aided cable resistance estimation approach is proposed. The ANN model is trained by simulation data. The trained ANN model can quickly and effectively map the current sharing ratios between the converters to the droop coefficients of the converters. In this way, the optimal droop coefficient combination that will yield the desired accurate current sharing ratio between the converters can be predicted by the trained ANN model. Subsequently, the optimal droop coefficient combination can be used in the estimation of the corresponding subsystem cable resistance by solving an equation set. The estimated cable resistance is compared with the simulated cable resistance and an excellent match is observed.

Full-bridge current doubler rectifier topology is used to restrict the ripple of output current and quicken the dynamic response. However, mass and power loss of filter composed of passive components are large. To optimize the output filter parameters, this paper adopts machine learning (ML) methods to train a support vector machine (SVM) model and an artificial neural network (ANN) model using data samples collected from simulation. SVM is used to judge the feasibility of filter design parameters, and the trained ANN serves as a dedicated surrogate model mapping from the design variables to the two optimization objectives (mass and power loss). After the ML aided filter optimization, the filter prototype based on the optimal design point is manufactured and tested on an experiment platform for the method validation.

Virtual prototyping of power electronic modules aims to allow rapid evaluation of potential designs without building and testing physical prototypes. Among the interests in thermal models of the virtual modules, process of compact thermal models needs effective methodology to fast generate small models describing the thermal performance of a potential design. This study chooses the Generalized Minimized Residual (GMRES) Algorithm to process thermal models due to its efficiency. Based on that, a machine learning aided surrogate model is proposed for the prediction of thermal performance since existing approaches take much time to determine the thermal response to a particular input power. This surrogate model is created by training a dedicated artificial neural network (ANN) on simulation data, after that this model can quickly map the module temperature and the power input in time domain. In the training process, cross-validation method is introduced to determine which neuron structure should be selected for the practical data generated by thermal equations. The test group is noted in cross-validation to give the prediction performance of structure candidates. To verify the proposed method, the resulting data of trained surrogate models are compared with the accurate simulation data after the ANN based cross-validation.