Identification of a Proton-Exchange Membrane Fuel Cell’s Model Parameters by Means of an Evolution Strategy

Abstract

This paper presents the parameter identification of an equivalent circuit-based proton exchange membrane fuel cell model. The model is represented by two electrical circuits, of which one reproduces the fuel cell's output voltage characteristic and the other one its thermal characteristic. The output voltage model includes activation, concentration, and ohmic losses, which describe the static properties, while the double layer charging effect, delays in fuel and oxygen supply, and other effects provide the model's dynamic properties. In addition, a novel thermal model of the studied Ballard's 1.2 kW Nexa fuel cell is proposed. The latter includes the thermal effects of the stack's fan which significantly improve the model's accuracy. The parameters of both, the electrical and thermal, equivalent circuits were estimated on the basis of experimental data by using an evolution strategy. The resulting parameters were validated by the measurement data obtained from the Nexa module. The comparison indicates a good agreement between the simulation and the experiment. In addition to simulations, the identified model is also suitable for usage in real-time fuel cell emulators. The emulator presented in this paper additionally proves the accuracy of the obtained model and the effectiveness of using an evolution strategy for identification of the fuel cell's parameters.

Generic Model of PEM Fuel Cells and Performance Analysis in Frequency Containment Period in Systems with Decreased Inertia

“…The accuracy of the developed model was tested against other models in the literature. First, by using (2)-(6), the stack voltage was calculated for each output current value from 0 to 45 A and compared to literature values from [20]. Fig.…”

Thermal and Electrical Parameter Identification of a Proton Exchange Membrane Fuel Cell Using Genetic Algorithm

“…Previous works dealing with PEMFC parameters identification have tested PSO [7,26], HADE [12] and EA [25]. HADE is an evolution in parameter identification that surpasses the PSO results and EA was tested to identify the thermal component of a PMFC.…”

“…One recent approach [25] used an equivalent electrical circuit model to represent a Nexa Ballard 1.2 kW PEMFC. This model simulated both the output voltage and the stack temperature.…”

Membrane‐hydration‐state detection in proton exchange membrane fuel cells using improved ambient‐condition‐based dynamic model

“…[33][34][35][36][37][38][39][40][41][42][43] These models use passive as well as active electrical components that are combined both in series and in parallel and can accurately emulate the voltage produced by PEMFC systems. [33][34][35][36][37][38][39][40][41][42][43] These models use passive as well as active electrical components that are combined both in series and in parallel and can accurately emulate the voltage produced by PEMFC systems.…”

“…The second type of PEMFC models includes electricalequivalent models. [33][34][35][36][37][38][39][40][41][42][43] These models use passive as well as active electrical components that are combined both in series and in parallel and can accurately emulate the voltage produced by PEMFC systems. A major drawback of these models is that they do not consider variations in ambient conditions.…”