As power system equipment gradually ages, the automated disassembly of transformers has become a critical area of research to enhance both efficiency and safety. This paper presents a transformer disassembly system designed for power systems, leveraging multimodal perception and collaborative processing. By integrating 2D images and 3D point cloud data captured by RGB-D cameras, the system enables the precise recognition and efficient disassembly of transformer covers and internal components through multimodal data fusion, deep learning models, and control technologies. The system employs an enhanced YOLOv8 model for positioning and identifying screw-fastened covers while also utilizing the STDC network for segmentation and cutting path planning of welded covers. In addition, the system captures 3D point cloud data of the transformer’s interior using multi-view RGB-D cameras and performs multimodal semantic segmentation and object detection via the ODIN model, facilitating the high-precision identification and cutting of complex components such as windings, studs, and silicon steel sheets. Experimental results show that the system achieves a recognition accuracy of 99% for both cover and internal component disassembly, with a disassembly success rate of 98%, demonstrating its high adaptability and safety in complex industrial environments.

This study aims to develop an efficient recovery solution for waste transformer insulating oil, addressing the challenge of incomplete separation of residual oil in existing recovery technologies. A multi-module integrated system is constructed, comprising a waste oil extraction module, a residual oil vaporization module, an exhaust gas treatment module, and an online monitoring module. By combining steps such as oil extraction, residual oil absorption, hot air circulation heating, and negative-pressure low-frequency induction heating, the complete recovery of waste oil is achieved. The recovery process incorporates oil–gas saturation monitoring and an oil–gas precipitation assessment algorithm based on neural networks to enable intelligent control, ensuring thorough recovery of residual oil from transformers. The proposed system and methods demonstrate excellent recovery efficiency and environmental protection effects during the pre-treatment of waste transformer oil. Experiments conducted on 50 discarded transformers showed an average recovery efficiency exceeding 99%, with 49 transformers exhibiting no damage to core components after the recovery process. From a theoretical perspective, this research introduces monitoring and control methods for transformer insulating oil recovery, providing significant support for the green processing and reutilization of discarded transformer insulating oil. From an application value perspective, the recovery process helps reduce environmental pollution and facilitates the disassembly of transformers. This enables better analysis of transformer operating characteristics, thereby enhancing the reliability and safety of power systems.

The escalating volume of waste electrical and electronic equipment (WEEE) poses a significant global environmental challenge. The disassembly of printed circuit boards (PCBs), a critical step for resource recovery, remains inefficient due to limitations in the adaptability and dexterity of existing automated systems. This paper proposes an intelligent disassembly system for PCB components that integrates a multimodal large language model (MLLM) with a multi-agent framework. The MLLM serves as the system’s cognitive core, enabling high-level visual-language understanding and task planning by converting images into semantic descriptions and generating disassembly strategies. A state-of-the-art object detection algorithm (YOLOv13) is incorporated to provide fine-grained component localization. This high-level intelligence is seamlessly connected to low-level execution through a multi-agent framework that orchestrates collaborative dual robotic arms. One arm controls a heater for precise solder melting, while the other performs fine “probing-grasping” actions guided by real-time force feedback. Experiments were conducted on 30 decommissioned smart electricity meter PCBs, evaluating the system on recognition rate, capture rate, melting rate, and time consumption for seven component types. Results demonstrate that the system achieved a 100% melting rate across all components and high recognition rates (90–100%), validating its strengths in perception and thermal control. However, the capture rate varied significantly, highlighting the grasping of small, low-profile components as the primary bottleneck. This research presents a significant step towards autonomous, non-destructive e-waste recycling by effectively combining high-level cognitive intelligence with low-level robotic control, while also clearly identifying key areas for future improvement.

This research introduces an advanced attention-driven model designed to optimize mobile shelf warehouse order-picking. Our model incorporates an enhanced masking mechanism and context-aware decoder, streamlining the order-picking process. In essence, our model presents an attention model based heuristic solution to the long-standing problem of order-picking optimization, leveraging the latest in attention-based deep learning techniques. The attention model is combined with Apriori and the Adaptive Large Neighborhood Search (ALNS) algorithm to solve the bilevel combinatorial optimization model for mobile shelves. Compared to existing methods, our innovative model shows superior performance, offering significant potential in warehousing solutions.