No abstract

No abstract

No abstract

It is the objective of this study to present the 3-dimensional comprehensive computational fuel cell dynamics (CFCD) model of a scaled-down single PEM cell featuring a nominal active surface of 0.0004 m2 for both conventional and serpentine flow field designs. The performance of these two designs are then analysed and compared keeping both anode and cathode inlet gases fully humidified. Then the parallel flow field was modelled under different inlet gas relative humidities (RHa/RHc) representing saturated, moderate, and dry conditions to observe the gain in cell performance. The higher pressure drop of the serpentine flow-field was demonstrated and the magnitude of which might seem to be negligible in the range of 100 Pascal. The simulation results show that a parallel flow-field design with appropriate humidity level can compete with the serpentine counterpart and gives considerably lower pressure drop across the cell. With a grid-independency analysis, it is suggested for the computational power available, the percentage error of important variables (species concentration and averaged current density) between the reference and finer mesh is negligible (< 3%) and the solution time is considerably less.

An ANFIS (Adaptive neuro-fuzzy inference system) controller which bases on Takagi-Sugenos method and combines the advantages of neural controller and fuzzy multi-variable controller has been studied and developed for the real-time control of EGR and VGT in a diesel engine. In the AVL-BOOST and Matlab/Simulink co-simulation environment, the control performance of ANFIS controller has been compared with those optimal control strategies based on a fuzzy logic controller. Results show the new controller can have more active control to EGR position and the optimal emission levels can be maintained.