Large-Signal Stability Analysis and Design of Finite-Time Controller for the Electric Vehicle DC Power System

Research output: Contribution to journalArticlepeer-review

20 Scopus citations

Abstract

In the electric vehicle dc power system (EVDCPS), a motor or converter can be regarded as a constant power load (CPL) when tightly controlled. Different from the resistive load, CPL may make the EVDCPS unstable. Thus, in addition to less overshoot and shorter setting time, the controller of the converter applied in EVDCPS should maintain the system stable. In this article, the finite-time controller (FTC) is proposed based on exact feedback linearization (EFL) to stabilize the EVDCPS feeding CPL. First, to handle the nonlinearity caused by the CPL, the affine nonlinear model of EVDCPS is transformed into the Brunovsky canonical form by using EFL. Then, the finite-time control is adopted and a PI compensator is designed to eliminate the steady-state error. Under the proposed control algorithm, rigorous proofs indicate that not only the closed system can be large-signal stable, but also the EVDCPS bus voltage can converge the reference in finite time. Furthermore, to reduce the design difficulty of FTC, the backstepping method is adopted to design a backstepping finite-time controller(BFTC). Finally, simulations and experiments of FTC and BFTC are carried out. Results indicate that both FTC and BFTC have large-signal stability, and although the robustness of BFTC is inferior to FTC, the setting time and overshoot are greatly improved compared with the traditional linear feedback controller.

Original languageEnglish
Pages (from-to)868-878
Number of pages11
JournalIEEE Transactions on Industry Applications
Volume58
Issue number1
DOIs
StatePublished - 2022

Keywords

  • Backstepping
  • PI compensator
  • constant power load (CPL)
  • dc-dc converter
  • finite-time control

Fingerprint

Dive into the research topics of 'Large-Signal Stability Analysis and Design of Finite-Time Controller for the Electric Vehicle DC Power System'. Together they form a unique fingerprint.

Cite this