To enhance chatter monitoring in polishing and grinding, we propose a model fusing deep separable convolutional neural networks (DCNN) and Gated Recurrent Unit. Signals are preprocessed via optimized variational mode decomposition (VMD) and wavelet denoising. The squeeze excitation deep separable convolutional neural networks gated recurrent units (SE-DCNN-GRU) model employs depth-wise convolution for multi-scale feature extraction. Experiments achieve 98.8 % accuracy for titanium alloy parts, offering a robust solution for complex conditions.

Bearing fault is a common reason for failure in industrial electric motors. The analysis of vibration signals measured on bearings in electric motors is challenging due to the presence of multiple cyclic components not related to a fault, which affects the analysis. This article aims to explain the structure of vibration signals measured on motor bearings using cyclostationary analysis, which investigates the content of cyclic components of the signal, which are crucial for interpreting faults.

In this paper, a new rotor rub-impact fault diagnosis method is proposed. The more sensitive fault features associated with the dynamics information are achieved based on the dynamic convolutional neural network. The experiment shows that the proposed method can obtain better diagnosis performance compared with traditional methods and that it has achieved the highest diagnosis accuracy of 98.5 % without noise and 98 % with noise. 

This study examines how adjusting the gap ratio (R_d) between diffuser and impeller blades in centrifugal compressors affects performance. Computer simulations reveal that increasing R_d improves airflow stability at critical stages, reducing energy losses caused by turbulent airflow. Higher R_d also lowers the minimum airflow requirement and enhances pressure recovery, leading to better efficiency. 

Aiming at the problems existing in the multi-objective trajectory planning of industrial robotic arms, a trajectory optimization method based on the improved sparrow search algorithm is studied and proposed. The multi-objective optimization function is constructed using the 3–5–3 polynomial interpolation to build the trajectory model.

A thermoelastic hydrodynamic lubrication (TEHL) model with interface micro-texture (IMT) evaluates load-bearing capacity. The model characterises the textured interface with micro-convex peaks (MCPs) by introducing an equivalent scale factor and an optimised distribution density function. A mathematical framework quantifies anti-scuffing load-bearing capacity (ASLBC) under steady-state TEHL, revealing coupled stiffness–damping in a micro-textured meshing interface (MTMI) under dynamic loading.

The ambition to save weight could be satisfied by adaptation of additive manufacturing. A mild steel powder was validated for dynamically loaded damper brackets for a high-speed train. For mechanical values and dynamic load, conformity could be confirmed. To meet the expectations in static and dynamic load cases, the team has used mature software tools for design and optimization. 

This study explores a new way to power small flying vehicles by using their wing movements to generate electricity. By combining a special energy-harvesting material with flexible wings made from silk protein, the design produces steady power and reduces reliance on batteries. Tests showed improved energy efficiency and stability. This approach could lead to lighter, more sustainable aircraft for future use. 

In Sect. 1, a coordinate system for tooth surface formation with cutter errors is established to derive the tooth surface equation. In Sect. 2, the tooth contact analysis model and load tooth contact analysis model are established, and the formulas of the tooth surface contact ellipse and main curvature direction angles are derived. In Sect. 3, the influences of the cutter errors on the contact ellipses, the tooth surface contact trace, and the loaded contact characteristics are analyzed.

In this paper, data-driven and genetic algorithms are combined to optimize the parameters of the Jonson–Cook model and improve the simulation accuracy. Experiments verify the accuracy of the model, and the influence of cutting data on the residual stress of the machined surface is discussed, which provides suggestions for the cutting process and tool design.

In this paper, an obstacle avoidance grasping path planning method based on monocular vision is proposed. This approach achieves 99.50 % recognition accuracy in complex scenes, and the smoothness and motion efficiency of the optimized trajectory are improved by about 60 % and 10 %, respectively.

In this paper, kurtosis-based improved particle swarm optimization (IPSO) is proposed to optimize the number of penalty factors and decomposition levels in variational mode decomposition (VMD). A method for fusing time domain metrics with entropy features and extracting their streamline features by principal component analysis (PCA) is also proposed. Experimental data validation shows that the method not only reduces the training time of the classification model but also improves the classification accuracy.

The gear system with backlash is a strongly nonlinear system, and the widely used harmonic balance method cannot meet the requirements. Additionally, the quasi-periodic and unstable periodic motion analytical solutions may be derived, and the principle of chaos generation can be further explained. Therefore, the effect of the generalized harmonic balance method DTE (dynamic transmission error) on system bifurcation and stability is studied.

A special deformation prediction model for titanium alloy in thin wall machining is established and calibrated by micromilling experiments. The model has been successfully evaluated by micromilling experiments. It shows that the dimensional errors of the thin wall made of Ti-6Al-4V are significantly reduced after on-machine radial cutting depth compensations. The relative errors have been reduced from 16.5 % to 3.9 %.