Gearboxes often have multiple simultaneous faults, which are hard to identify when speeds change. We tested a new computer method that converts vibration data into simple network graphs. Using 64 public recordings, it reached 83 % average accuracy in finding seven fault types. It beat random graphs but scored lower than standard wave-based deep learning. This shows the promise and current limits of graph methods for machinery health monitoring.

This research studies the working equipment of hydraulic excavators, which bears heavy loads during digging operations. We collected real digging movement data from field tests and analyzed the structural performance under different working loads. We found the stress and deformation distribution of key components under practical digging trajectories. The results help engineers better understand structural load features and support safer and more reliable design of excavator working parts.

Special threaded connections must stay sealed and strong under combined heat, pulling, and pressure, but testing every condition is slow and costly. We built computer models of realistic machining and loading combinations, then trained machine learning to predict key measures like contact length and stress. Predictions matched simulations closely, and thread taper had the greatest influence. This helps manufacturers choose better machining settings and inspect dimensions more reliably.

Gear shafts in ships and machines work under heavy loads. Striking a metal surface with rapid ultrasonic pulses makes it harder, smoother and longer-lasting by squeezing its outer layer. But warmth during operation may weaken this benefit. We treated a steel gear shaft this way, then kept it warm. The surface stayed hard and smooth; heat only slightly eased the squeezing, while hardening remained intact. Computer predictions matched our tests, supporting safe use of this method in warm service.

High-speed magnetic levitation trains require a stable gap between the vehicle and guideway for safe and smooth operation. We developed a data-driven prediction method using detailed computer simulations to estimate this gap quickly from track irregularities. The method reproduces key response patterns more accurately than several existing approaches while greatly reducing calculation time. This can support faster repeated assessment of train performance under different operating conditions.

This paper proposes a parallelogram-based adaptive grasping device to overcome the limitations of conventional rigid grippers. It integrates adaptive parallelogram units and a 2-RRR planar parallel driving mechanism. Screw theory and kinematic analysis validate its performance, while a bevel gear set minimizes driving motors. Grasping tests confirm its favorable dimensional adaptability, providing theoretical and engineering references for adaptive grasping equipment design.

Ice and strong winds can seriously damage transmission lines and disrupt power supply. This review compares major methods used to understand and predict these effects, from physical and numerical models to statistical and data-driven approaches. It shows that no single method is sufficient and highlights the value of combining complementary methods to improve risk assessment, resilience, and future decision-making.

Autonomous vehicles must detect nearby road users quickly and reliably, even when objects are partly hidden. We developed a lightweight camera-based method that adapts its processing to scene difficulty, combines information across video frames, and learns from laser-sensor data during training. It reached 38.7% on the benchmark detection measure and processed 26.2 frames per second without laser sensors during use, showing strong potential for fast, lower-cost vehicle perception.

Aiming at stick failures under hard‑soil excavation, this paper presents a mechanism‑structure collaborative optimization for a 20‑ton excavator. With test‑identified hazardous conditions and Kriging surrogate models, NSGA‑II‑based two‑level optimization boosts digging force by 9 %, cuts peak stress by 7 % and reduces stick weight by 9 %. It provides references for excavator‑attachment multidisciplinary optimization.

Robotic grasping in complex environments is challenging due to variations in object positions, orientations, and backgrounds. This study proposes a vision-based method for collaborative robots that integrates object detection, target mask generation, and grasp position estimation to improve grasping reliability. Experiments on a robot platform demonstrate effective target-specific grasping in unstructured multi-object scenes, showing its potential for intelligent manufacturing applications.