Thin-walled structures have been widely used in various parts of vehicle subsystems because of their high-efficiency impact energy absorption and lightweight characteristics. However, the impact deformation mode of conventional thin-walled structures is unstable and the energy absorption efficiency is low. Therefore, a series of novel bionic conch structures (BCS) are proposed to find a more excellent crashworthiness design in this study. First, the finite element simulation model of BCS verified by experiments is established. Then, the energy absorption characteristics of bionic conch structures, and conventional single-cell and multi-cell tubes under axial loading are compared by employing finite element simulation. The results show that the thin-walled structures inspired by conchs have a higher energy absorption efficiency than the other two structures with the same mass. In addition, the influence of main design parameters (wall thickness, inner and outer ring diameter, and the number of inner and outer panels) on the crashworthiness of BCS is studied through parameter design and factor significance analysis. Finally, the optimal geometric configuration is found by combining the approximation model and multi-objective particle swarm optimization, and the crashworthiness of BCS is further optimized. The bionic crashworthiness design and optimization framework proposed in this study can also provide a reference for other engineering protective structures.

To study the influence of the geometric structure of the probe coil on the electromagnetic characteristics of the eddy current probe in the process of eddy current testing, based on the principle of eddy current testing, different probe coil models were established using finite element software. These geometric structure parameters include the difference between the inner and outer radius, thickness, and equivalent radius. The magnetic field distribution around the probe is simulated and analyzed under different parameters, and the detection performance of the probe is judged in combination with the change rate of the magnetic field around the probe coil. The simulation results show that at a closer position, increasing the difference between the inner and outer radii, reducing the thickness, and reducing the equivalent radius are beneficial to improve the resolution of the probe coil. At a far position, reducing the difference between the inner and outer radii, increasing the thickness, and reducing the equivalent radius are beneficial to improve the resolution of the probe coil. At the same time, the accuracy of the simulation data is verified by comparing the theoretical values with the simulated values under different conditions. Therefore, the obtained conclusions can provide a reference and basis for the optimal design of the probe structure.

In order to explore the failure mechanism of a reciprocating compressor system with clearance fault, we implemented a computational framework whereby a simulation model of the mechanism is established using ADAMS software in this paper, and a typical reciprocating compressor model is introduced to validate the design model. In this work, the joint clearance faults between the crankshaft and linkage, between the linkage and crosshead, and in both locations are taken into account computationally. These faults are one of the major causes of vibration. Through dynamic calculation and analysis of a system with clearance fault, the simulated results show that these clearance faults directly influence the vibration. The larger the gap size, the more severe the vibration and the higher the amplitude of the vibration. Furthermore, the clearance number also affects the vibration greatly.

To address the problem of insufficient real-world data on planetary gearboxes, which makes it difficult to diagnose faults using deep learning methods, it is possible to obtain sufficient simulation fault data through dynamic simulation models and then reduce the difference between simulation data and real data using transfer learning methods, thereby applying diagnostic knowledge from simulation data to real planetary gearboxes. However, the label space of real data may be a subset of the label space of simulation data. In this case, existing transfer learning methods are susceptible to interference from outlier label spaces in simulation data, resulting in mismatching. To address this issue, this paper introduces multiple domain classifiers and a weighted learning scheme on the basis of existing domain adversarial transfer learning methods to evaluate the transferability of simulation data and adaptively measure their contribution to label predictor and domain classifiers, filter the interference of unrelated categories of simulation data, and achieve accurate matching of real data. Finally, partial transfer experiments are conducted to verify the effectiveness of the proposed method, and the experimental results show that the diagnostic accuracy of this method is higher than existing transfer learning methods.

Quantization is a widely adopted technique in model deployment as it offers a favorable trade-off between computational overhead and performance loss. Integer-arithmetic-only quantization is an important approach in quantization and holds great significance for hardware deployment with limited resources. However, existing methods represented by IAO face challenges when balancing hardware efficiency and accuracy. They suffer from significant accuracy loss and the multipliers in element-wise layers are not constrained to be power-of-2. These issues limit the utilization of hardware resources. In this paper, we explore integer-arithmetic-only quantization and introduce two techniques: re-quantization and re-scale. Our approach ensures that inference in an 8-bit quantized network involves only 8-bit multiply-accumulate operations and bit-shifting. We compare our method with previous integer-arithmetic-only approaches and demonstrate that our approach not only accelerates inference speed and reduces resource consumption but also achieves minimal performance degradation. For the commonly used ResNet-50, our int8 model exhibits only a 0.5% drop in Top-1 accuracy, significantly outperforming previous integer-arithmetic-only methods. Moreover, the proposed method frees up digital signal processing resources and improve parallelism, achieving an improvement of approximately 27% in frames per second and reducing inference time by 27.09 ms (19%). This showcases the practical value and effectiveness of our proposed method in improving the overall efficiency of model inference.

This article takes the linear inclined vibrating screen (TLIVS) as the research object, establishes a spatial 6-degree-of-freedom dynamic model of TLIVS, and verifies the correctness of the established model through comparison between real experiments and simulation experiments. Steady-state response of TLIVS is solved by numerical analysis method, The influence of each degree of freedom acting separately on the screening process of TLIVS was explored using the discrete element method. The results indicate that x-direction translation movement mainly promotes the dispersion of the mixture, and y-direction translation movement mainly increases the velocity of the mixture towards the outlet. z-direction translation movement mainly causes the mixture to stratify. Rotation in the x-direction accumulates the mixture in the middle of the screen surface, rotation in the y-direction helps to disperse the material at the inlet, and rotation in the z-direction may cause blockage of the mixture.