A Lightweight and Smart Deep Kernel Network Learning System for Energy Management and Control in Electrical Ships

The adoption of completely electric power ships is a relatively new emergent technology, owing to the rising ecological impacts of ship emissions and tightening preventive regulations. Fuel cell systems are an exciting development, which renders them an appealing option for sailors to employ as their main source of energy. The main focus of this research attempt is to facilitate the build a new energy management system with smart control designs that can handle the load requirements of shipboard applications. For this purpose, the proposed work leveraged a combination of renewable sources, such as FCs and battery storage, to provide shipboard electrical requirements. Then, the Artificial Fish Swarm Optimized Tracker (AFSOT) is adopted to maximize the energy production from FCs for satisfying the need of shipboards. The unique Fused Ant Lion—Crow Optimized Controller (FAL-COC) is deployed for enhancing the controlling performance of a non-isolated high gain interleaved converters. With proper controlling operation, the voltage regulation and high efficiency are greatly improved in the proposed framework. Moreover, High Gain Z-Source integrated Boost (HGZB), a lightweight inverter topology is being used to reduce inrush current with less complexity. Specifically, the smart Deep Kernel Network Learning (DKNL) methodology is developed to enable an effective and successful energy management in ship applications. During performance study, multiple parameters are used to validate the results and consequences of proposed energy management and controlling methods.