This study uses a relatively novel optimization system to optimize the structure of a self-excited oscillating nozzle. The study identifies inlet radius, lower-nozzle diameter, and cavity length as key variables (combined impact >87 %); achieves a 77.39 % improvement in overall cleaning efficiency through global optimization; and explains the flow field evolution mechanism.

Self-balancing wheelchairs can help people move in tight spaces, but they must stay upright when starting from a large tilt, carrying different users, or facing sudden pushes. We developed and tested a two-stage control approach that first brings the chair back safely from larger tilts and then keeps it steady near upright. Simulations and prototype tests showed faster recovery, better stability under disturbances, and stronger tolerance to user-weight changes, supporting safer designs.

Speed skating training demands precise movement, but traditional coaching relies on subjective experience and existing tools are bulky or inflexible. We created a flexible cable-driven robot with enhanced control, cutting tracking errors significantly to enable precise training and inform smart sports equipment development.

Traditional mathematical methods for calculating machining errors before production are often time-consuming. To solve this problem, we developed a data-driven intelligent method that learns from existing data to realize real-time error prediction. Our method achieves over 99 % prediction accuracy and runs significantly faster than conventional approaches, helping factories manufacture high-precision, high-quality parts more efficiently.

We developed an improved control method for UR10 robot arms, which typically face issues with unpredictable forces and friction. By combining an online-learning neural network with a robust sliding-mode controller, our approach enables more accurate trajectory tracking and eliminates jitter. 

High-speed train collisions with obstacles can cause derailment. To address this concern, we conducted a study to understand how different collision angles and speeds affect train stability. Our key findings show that as the angle increased from 0 to 45°, the speed at which derailment occurred dropped significantly. These results help identify safety speed limits and guide infrastructure design to reduce collision risks, making high-speed rail travel safer for everyone.

This study proposes a multi-objective optimization approach tailored for a heavy-duty wave compensation parallel platform, which integrates an artificial neural network (ANN) surrogate model with an enhanced non-dominated sorting genetic algorithm III (NSGA-III).

We developed a wear model for helical gears with surface modifications, coupling wear with dynamic contact forces and a dynamic wear coefficient. Unlike traditional constant‑rate models, our approach captures real‑time changes in contact stress. Results show wear has little effect on gear stiffness, but its distribution varies strongly with modification type, which also alters vibration evolution. This work aids selection of surface designs to improve gear reliability and reduce noise.