Artificial intelligence-driven methodological insights into wireless charging and sustainable energy management schemes: A survey

Wireless rechargeable sensor networks (WRSNs) consist of sensors equipped with fast-rechargeable batteries and wireless power receivers, where energy storage plays a crucial role in bridging the intermittent nature of wireless power transfer (WPT) with the continuous energy demands of the sensors. The mobile charger (MC) enables network functionality through wireless power transmission, effectively addressing the charging requirements of these networks. This survey seeks to illuminate various aspects of wireless charging schemes regarding management strategies and technical methodologies, particularly in the context of WPT and energy consumption management from a methodological standpoint. By presenting a comprehensive, methodology-driven analysis of wireless charging schemes for WRSNs, this survey makes a significant contribution to the field. Furthermore, this survey aims to provide artificial intelligence (AI)-driven methodological insights into various facets of wireless charging schemes within the framework of sustainable energy management. It examines innovative management strategies and technical methodologies that utilize artificial intelligence to enhance the efficiency of WPT and energy consumption. Each wireless charging scheme is scrutinized through three distinct lenses: technical methodology, charge methodology, and evaluation methodology, considering the specific design parameters relevant to each approach. The technical methodology explores the capabilities and limitations of each scheme, detail the analytical parameters involved, and define the operational objectives. The charge methodology focuses on charging characteristics, including management patterns, operational modes, and categorization of charging processes. The evaluation methodology provides a holistic overview of how these schemes are assessed, encompassing evaluation metrics, simulation configurations, and the simulation environment. To better understand the landscape of these methodologies, this survey categorizes the reviewed schemes into five distinct classes: heuristic-based, metaheuristic-based, fuzzy-based, hybrid approaches that integrate both metaheuristic and fuzzy logic techniques, and reinforcement learning-based techniques. This categorization is grounded in the aforementioned three innovative aspects of technical, charge, and evaluation methodologies. Importantly, this survey aims to offer an exhaustive overview of the capabilities, limitations, and emerging research directions concerning these schemes from a methodology-driven perspective, particularly in the realm of WRSNs. Additionally, it seeks to pave the way for future research by identifying new avenues for exploration, filling the gap left by the lack of comprehensive surveys in this crucial field.