Model predictive control for PEMFC based on least square support vector machine

Lu, Jun, and Zahedi, Ahmad (2012) Model predictive control for PEMFC based on least square support vector machine. In: Asia Pacific Power and Energy Engineering Conference (APPEEC), 2012, pp. 1-4. From: Power and Energy Engineering Conference (APPEEC), 2012, 27-29 March 2012., Shanghai, China.

[img] PDF (Published Version) - Published Version
Restricted to Repository staff only

View at Publisher Website: http://dx.doi.org/10.1109/APPEEC.2012.63...
 
2
8


Abstract

The modelling and control of PEMFC possess great challenges due to PEMFC system's inherent nonlinearities and time-varying characteristics. The objective of this paper is to propose a novel model predictive control (MPC) strategy based on the least square support vector machine (LSSVM). First, a set of LSSVM models are generated and each model represents one PEMFC performance output. By mapping PEMFC performance outputs as a function of various operation conditions, LSSVM models disregard complex internal details and thus provide low calculation burden for the control algorithm. Then model predictive control is employed by using Model Predictive Control Toolbox in the MATLAB/SIMULINK environment. The LSSVM models and the model predictive controller are simulated and the results demonstrate their effectiveness. The nominal value of the oxygen excess ratio and the stack voltage are able to be maintained during abrupt changes in the stack current.

Item ID: 25058
Item Type: Conference Item (Refereed Research Paper - E1)
Keywords: least square support vector machine, model predictive control, proton exchange membrane fuel cell
ISBN: 978-1-4577-0546-5
ISSN: 2157-4839
Date Deposited: 20 Feb 2013 09:30
FoR Codes: 09 ENGINEERING > 0906 Electrical and Electronic Engineering > 090608 Renewable Power and Energy Systems Engineering (excl Solar Cells) @ 100%
SEO Codes: 85 ENERGY > 8505 Renewable Energy > 850599 Renewable Energy not elsewhere classified @ 100%
Downloads: Total: 8
More Statistics

Actions (Repository Staff Only)

Item Control Page Item Control Page