Modeling and control of Proton Exchange Membrane (PEM) fuel cell system.

摘要

This dissertation presents the design, implementation and application of soft computing methodologies to Proton Exchange Membrane (PEM) Fuel Cell systems. In the first part of the research work, two distinct approaches for the modeling and prediction of a commercial PEM fuel cell system are presented. Several Simulink models are constructed from the electrochemical models of the PEM fuel cells. The models have been simulated in three dimension (3-D) space to provide the visual understanding of fuel cell behaviors. In addition, two optimal predictive models, based on back-propagation (BP) and radial basis function (RBF) neural networks are developed. Experimental data as well as pre-processing data are utilized to determine the accuracy and speed of the proposed prediction algorithms. Extensive simulation results are presented to demonstrate the effectiveness of the proposed method on prediction of nonlinear input-output mapping.;In the second part of the study, the design and implementation of several fuzzy logic controllers (FLCs) and neuro-fuzzy controllers as well as classical controllers are carried out. The proposed real-time controller design is based on the integration of sensory information, Labview programming, mathematical calculation, and expert knowledge of the process to yield optimum output power performance under variable load condition. The implementations of the proposed controllers are carried out for a commercial PEM fuel system at FAU Fuel Cell Laboratory. The performance of the proposed controllers pertaining to the oxygen (O2) flow rate optimization as well as the actual fuel cell output power under a variable load bank are compared and investigated. It was found the Fuzzy Logic Controller design provide a simple and effective approach for the implementation of the fuel cell systems.