Prediction-Based MIMO Control of a Multiport Magnetic-Coupled Energy Router

Multiport power converters are important for energy routing in renewable energy applications. Particularly, magnetic-coupled multiport converters are especially advantageous in terms of galvanic isolation and reduced number of magnetic components. They also enable highly integrated and modularized circuit designs. Meanwhile, their multiple-input and multiple-output (MIMO) nature complicates the control, particularly with a possibly large number of ports for routing purposes. Unlike the existing single-input and single-output (SISO) controls, this article develops a continuous-control-set model predictive control (MPC)-based MIMO control scheme that improves the dynamic performance of the multiport power converters, minimizes interference among different ports, and simplifies the controller design process. The aspects of state-space MIMO modeling, optimization problem definition and solution, and necessary modifications are systematically discussed for practical real-time implementation, especially on low-cost microcontroller units (MCUs). Finally, the developed MPC-based MIMO control is experimentally validated by a prototype eight-port energy router (800 W maximum power) exhibiting superior dynamic responses and a high system efficiency above 94% over a wide power range.