Neural-PID Control of Air Pressure in Fuel Cells

Abstract

Air pressure in proton exchange membrane fuel cells (PEMFC) directly affects fuel cells' electrochemistry-reactivity performance. It not only related to air flow and air temperature, but also affected by PEMFC output power. In order to make it in a proper value, the paper designed air pressure servo control system which given value was predicted in advance by Elman ANN according to PEMFC output power, air flow and air temperature. Its controller chose neural-PID, and control model adopted identifying ways basing on Elman ANN, and its control variable was fuel cell's outlet air flow. Simulation result shows that air pressure was adjusted in time by neural-PID with output power changes.

A review of adaptive neural control applied to proton exchange membrane fuel cell systems

“…10, PID parameters can also be directly estimated by a neural model. A direct approach was undertaken by Guoai et al [69] to develop a power control of a 60kW PEMFC via the air pressure. The controller is a self-tuned neural-PID based on Elmann-type NN.…”

A nonlinear control strategy for fuel delivery in PEM fuel cells considering nitrogen permeation

“…A neural proportion integration differentiation control strategy was studied in Ref. [5] for the air pressures of the fuel cells. Model predictive control (MPC) strategies are proposed in Ref.…”