Recursive Estimation-Based Maximum Power Extraction Technique for a Fuel Cell Power Source Used in Vehicular Applications

Abstract

Extraction of maximum power from a fuel cell (FC) power source (PS) is essential for its optimum and economical utilization. However, the maximum extractable power from an FCPS varies dynamically during the fuel cell operation for varying load current requirements as the system parameters are also changing. One such example is the use of an FCPS in vehicular applications, where the power requirement varies dynamically during the drive cycle. This makes maximum power extraction a challenging task. As the load varies, the equivalent resistance appearing across the FC varies too. The maximum power point (MPP) appears prominently on the power versus curve. This paper presents a novel MPP tracking (MPPT) scheme using nonlinear curve fitting and recursive least-squares estimation (RLSE). A current controlled Cuk converter is used due to its low ripple feature. RLSE determines the MPP online, which is used as a reference for the control of the Cuk converter. The performance of this MPPT scheme for a typical vehicular drive cycle is compared with the popular “Perturb and Observe” and Incremental conductance methods. All the analytical and simulation results are included. Experimental results are also presented to validate the proposed scheme for the drive cycle load profile.