Fuel Consumption Optimization in Solid Oxide Fuel Cell Based Hybrid Electric Tractor Using Multihead Cross Attention and Sea‐Horse Optimization

ABSTRACT In the research, improving the efficiency and reducing fuel consumption in Solid Oxide Fuel Cell‐based Hybrid Electric Tractors (SOFC‐HET) by employing a hybrid approach is proposed. The proposed strategy is the Evolved Multihead Cross Attention based Bidirectional long short‐term memory Network (EMCABN). For enhancing the precision of EMCABN predictions, Sea‐Horse Optimizer (SHO) is introduced. EMCABN is employed in the prediction of the tractor power demand and the operational status of the fuel cell. SHO directly focuses on optimizing the control parameters associated with power distribution, fuel flow, air flow, and temperature to minimize fuel consumption effectively. The input parameters to be optimized are SOFC nominal power, battery capacity, hydrogen consumption, cost of electricity, and so forth. For validation, the system is tested under the MATLAB platform and compared with other approaches. Under the proposed technique, the hydrogen tank capacity during plowing with initial State of Charge (SOC) levels of 40%, 65%, and 90% is 6.457 kg.