Power buffer with model predictive control for stability of vehicular power systems with constant power loads

Xinan Zhang, D. Mahinda Vilathgamuwa, King Jet Tseng, Bikramjit Singh Bhangu, Chandana Jayampathi Gajanayake

Research output: Contribution to journalArticle

31 Citations (Scopus)


Electrification of vehicular systems has gained increased momentum in recent years with particular attention to constant power loads (CPLs). Since a CPL potentially threatens system stability, stability analysis of hybrid electric vehicle with CPLs becomes necessary. A new power buffer configuration with battery is introduced to mitigate the effect of instability caused by CPLs. Model predictive control (MPC) is applied to regulate the power buffer to decouple source and load dynamics. Moreover, MPC provides an optimal tradeoff between modification of load impedance, variation of dc-link voltage and battery current ripples. This is particularly important during transients or starting of system faults, since battery response is not very fast. Optimal tradeoff becomes even more significant when considering low-cost power buffer without battery. This paper analyzes system models for both voltage swell and voltage dip faults. Furthermore, a dual mode MPC algorithm is implemented in real time offering improved stability. A comprehensive set of experimental results is included to verify the efficacy of the proposed power buffer.

Original languageEnglish
Article number6380631
Pages (from-to)5804-5812
Number of pages9
JournalIEEE Transactions on Power Electronics
Issue number12
Publication statusPublished - 24 Jun 2013
Externally publishedYes


Cite this