Fast Nonlinear Model Predictive Control for the Energy Management of Hybrid Energy Storage System in Wave Energy Converters

Abstract

The integration of a hybrid energy storage system (HESS) into a wave energy converter (WEC) helps achieve smoother power output, provided that an effective energy management strategy (EMS) is employed. This article proposes a fast nonlinear model predictive control (NMPC)-based EMS that balances multiple objectives while maintaining an acceptable computational burden. First, a multiobjective cost function is formulated to optimize system efficiency, battery life, and supercapacitor voltage regulation. The numerical solution to the constrained optimal control problem (OCP) is confined to an adaptive admissible control set. To efficiently solve the OCP, a fast forward dynamic programming (FDP) algorithm is designed. The curse of dimensionality in dynamic programming is mitigated using a MB strategy and state-space approximation. Furthermore, a “smart-select” technique, employing flexible resolution and early termination, is introduced to prune unnecessary search paths. Finally, comparative case studies are conducted under various load conditions, confirming the superior performance of the proposed EMS in reducing energy loss, battery ampere-hour throughput, and battery rms current. Moreover, the significantly improved computational efficiency on an embedded controller further demonstrates the effectiveness of the fast FDP algorithm.