Castulo Aaron De La O

In this paper, a novel approach for real-time, probabilistic digital twinning is proposed for the online diagnostic analysis of power electronic converters. Under this approach, a digital twin of a power converter is defined as a real-time probabilistic simulation model with stochastic (random) variables, the model divided up in terms of control layers for converter subsystems, and developed using generalized Polynomial Chaos Expansion (PCE). Real-time executed solvers of these divided probabilistic digital twin models are embedded into power converter controllers running on FPGA devices. Using a monitoring system, these digital twins and the power converter are compared by the power converter controller to determine if the converter is operating within probable behavior. Knowing the large computational cost of probabilistic modeling, the resource usage and timing of this real-time digital twin approach on modern FPGAs is reported for common converter topologies, showing the approach is feasible and able to perform probabilistic real-time simulation of smaller power converters in perspective of control layers using $\leq 2\mu s$ time steps, with as low as $70ns$ steps. To highlight the potential of the proposed approach, a case study is presented using a probabilistic digital twin in the controller of a pair of converters under HIL testing.