Deep Learning and NLP-based Trend Analysis : in Actuators and Power Electronics

Actuators and power electronics are indispensable elements of contemporary control systems, supporting high-precision operation, optimized energy utilization, and advanced automation capabilities. This study explores research trends and thematic developments in these domains over the past two decades (2005–2024). An analysis was conducted on 1,840 peer-reviewed abstracts sourced from the Web of Science database, utilizing BERTopic modeling, which combines transformer-based sentence embeddings with UMAP for dimensionality reduction and HDBSCAN clustering. The methodology further incorporated class-based TF-IDF, intertopic distance mapping, and hierarchical clustering to elucidate structural topic formations. Results demonstrate a consistent rise in scholarly output, with notable acceleration after 2015. Research between 2005 and 2014 was predominantly concentrated on established areas such as piezoelectric actuators, adaptive control, and hydraulic systems. Conversely, the 2015–2024 phase exhibited an expansion into emerging themes like advanced materials, robotic mechanisms, fault-tolerant systems, and networked actuator control via communication protocols. The structural analysis highlighted a progression from a cohesive to a more distributed and specialized array of topics. This study provides a robust, data-informed perspective on the progression in complexity and breadth within actuator and power electronics research. The insights derived hold significance for researchers, engineers, and policymakers committed to fostering advanced, sustainable solutions in next-generation industrial technologies.