A Hybrid Adaptive Control Architecture for Grid-Connected Inverter With Optimal Policy Generation

A nonlinear optimization-based hybrid control architecture for grid-connected inverters is presented in this article. The proposed control approach augments the existing inverter control with an identification-based adaptive control architecture. Grid dynamics are first represented as energy signals at the inverter terminals using the concept of an energy function. The nonlinear optimization function will act on this dynamic energy signal and derive an optimal control gain for the augmented hybrid control architecture. The optimization function derived in this work considers both the state energy and controller effect. The proposed architecture developed considering the nonlinearities in the system ensures excellent performance during high-frequency grid dynamics. Based on validation on a real power grid, it was observed that the proposed architecture provides better control accuracy (more than 20%) when compared to static PI controller acting alone.