Date of Award

2008

Degree Type

Thesis

Department

Chemical and Biomedical Engineering

First Advisor

Ungarala, Sirdhar

Subject Headings

Proton exchange membrane fuel cells, Fuel cells, Kalman filtering, Filters and filtration

Abstract

Research on alternative and renewable energy sources which are amicable to the environment has gained momentum because of the growing concern about the tremendous increase in the concentration of toxic and green house gases and scarcity of the fossil fuels. Among the available renewable sources, fuel cell technology has received a high research attention due to their high efficiency and superior reliability. Among the various fuel cells available, Polymer electrolyte membrane fuel cell is promising source for both stationary and mobile applications because of its high efficiency and low operating temperatures. The performance of the fuel cell depends on the partial pressure of the hydrogen and oxygen, temperature of the stack and membrane humidity. A major obstacle in achieving active control of membrane water content and reactant supply is lack of reliable measurements of partial pressure of the gases and membrane humidity which motivates the use of estimators for estimating the partial pressure of the reactants. This thesis investigates the use nonlinear estimators such as sequential Monte Carlo and unscented Kalman filter to the estimate the partial pressure of hydrogen and oxygen and temperature. The performance of the two filters is studied for cases of poor filter initialization, plant-model mismatch and multiple load variations by calculating the mean square error. The performance of unscented Kalman filter was better than the sequential Monte Carlo which was not anticipated

Share

COinS