Self-Healing Predictive Control of Battery System in Naval Power System with Pulsed Power Loads

Abstract

This article presents a model predictive self-healing control (MPSC) scheme for battery system interfaced dual active bridge (DAB) converter in navy ship power system (NSPS) with pulsed power loads (PPLs). The voltage and frequency of NSPS are vulnerable to PPLs energization. A properly controlled battery system with fast dynamic response can mitigate this vulnerability. Model predictive control (MPC) is a potential solution for the battery system interfaced DAB converter to achieve fast dynamic response and mitigate disturbances imposed to the NSPS by PPLs. However, conventional MPC framework suffers from current prediction error due to the pulsating AC-link inductor's voltage profile in DAB converter. This article proposes a self-healing control loop that utilizes the feasible range of power transfer in conjunction with the AC-link inductor's voltage profile. The proposed method can validate and autonomously correct the predicted current and phase shift in DAB converter interfaced a battery system. The proposed control scheme on DAB converter prevents voltage and frequency collapse in a hybrid AC/DC NSPS particularly during the PPL energization. The system stability is studied based on Lyapunov stability analysis. The theoretical concepts are validated by several case studies implemented on a hardware-in-the-loop (HIL) testbed of a NSPS. The case studies demonstrate voltage and frequency regulation of the NSPS with fast dynamic response during PPLs energization. The proposed MPSC performance is compared with proportional-integral (PI) based control for DAB in NSPS with PPLs.