Comparison of Decentralized ADMM Optimization Algorithms for Power Allocation in Modular Fuel Cell Vehicles

The advanced modular powertrains are envisioned as primary part of future hybrid fuel cell vehicles (FCVs). existing papers in the literature solely cope with hardware side modularity, while software is also vital to capitalize on total capacity these powertrains. Driven by this motivation, article puts forward a comparative study two novel decentralized convex optimization frameworks based alternating direction method multipliers (ADMM) solve multi-objective power allocation strategy (PAS) problem FCV (MFCV). MFCV composed (FC) stacks and battery pack. Despite centralized strategies for such system, manuscript proposes PASs (Dec-PASs) Consensus ADMM (C-ADMM) Proximal Jacobian (PJ-ADMM) bridge gap regarding appreciation modularity terms. Herein, after formulating central PAS problem, principle utilizing algorithms presented detail. Subsequently, performance proposed Dec-PASs examined through several numerical simulations well experiments developed small-scale test bench. obtained results illustrate that decomposition into forms enables solving complex faster provides flexibility. Furthermore, can fault malfunction thus augment durability robustness powertrain systems.