The lithium-ion battery is widely used in the power system of pure electric vehicles and hybrid electric vehicles due to its high energy density. However, the chemical and electrochemical reactions generate a lot of heat. If the heat is not transferred through some refrigeration methods in time, it will lead to a rapid rise in the temperature of the battery. In this paper, an electric–thermal coupling model of a cylindrical Panasonic 21700 battery was proposed by using offline parameter identification method. Based on this model, a battery thermal management system using a heat pipe was established. The experimental results show that the model can simulate the actual performance of battery well. When the ambient temperature is 25 °C, the battery parameters change little and battery performance is better. The heat pipe battery thermal management system performs better than the non-heat pipe battery system in the discharge process, and can control the battery temperature well at low and high temperatures. Changing the refrigerant temperature can achieve a better thermal management effect under suitable ambient temperature conditions.

Micro-tubes have small diameters and thin wall thicknesses. When using double-layer gas-assisted extrusion (DGAE) technology to process micro-tubes, due to the influence of flow resistance, airflow from the inner gas-assisted layer cannot flow into the atmosphere through the lumen. Over time, it will inflate or even fracture the micro-tubes intermittently and periodically. To solve this problem, a new double-layer micro-tube gas-assisted extrusion die was designed in this study. Its mandrel has an independent airway leading to the lumen of the extrudate, with which the gas flow into the lumen of the extrudate can be regulated by employing forced exhaust. Using the new die, we carried out extrusion experiments and numerical calculations. The results show a significant positive correlation between micro-tube deformation and gas flow rate in the lumen of a micro-tube. Without considering the refrigerant distortion of the microtube, the flow rate of forced exhaust should be set equal to that of the gas from the inner gas-assisted layer flow into the micro-tube lumen. By doing this, the problem of the micro-tube being inflated can be eliminated without causing other problems.

In the process of double-layer gas-assiste extrusion of a plastic micro-tube, the tugging effect caused by the pressure difference of the gas cushion layer inside the die has a great influence on the external dimensions of the micro-tube. Therefore, this study establishes a two-phase extrusion model based on compressible gas and incompressible melt. Ansys Polyflow finite element software was used to numerically simulate the extrusion process of the melt to analyze the effect of the gas cushion layer pressure difference on the micro-tube deformation. The research shows that the shrinkage rate of the micro-tube increases with increasing pressure of the outer cushion layer, and the degree of tube wall migration increases, too. In the process of extrusion, the first normal stress difference at the entrance of the gas cushion layer shows a significant effect on the melt velocity field distribution and the extruded micro-tube cross-sectional deformation.

Unsupervised anomaly detection is a crucial technology in the field of intelligent manufacturing. The conventional single-class paradigm of ‘one-class-one-model’ faces challenges of surging training and deployment costs in multiclass scenarios. Directly applying single-class methods to multiclass scenarios leads to significant performance degradation, as the model learns diverse feature information during the training phase, which further amplifies the ‘identity shortcut’ issue. To address this challenge, this paper proposes a multiclass anomaly detection model with directional feature perturbation (MAD–DFP). We regard the ‘identity shortcut’ issue as an overgeneralization of the reconstruction model to anomaly features and propose targeted strategies from the perspective of feature perturbation. In particular, MAD–DFP precisely locates key feature regions under the guidance of attention and perturbs features, thereby suppressing the model’s overgeneralization problem. Meanwhile, an anomaly map fusion module is proposed to adaptively fuse multilevel feature maps, avoiding information loss caused by simple averaging. Extensive experiments on MVTec-AD, VisA, BTAD, and MPDD reveal that MAD–DFP achieves optimal image-level and pixel-level performance in multiclass scenarios. Notably, the model exhibits a slight performance gap between single-class and multiclass settings, demonstrating strong application potential.

Gas-assisted extrusion is an effective method for improving the deformation behavior of polymer catheters during extrusion. However, the underlying mechanisms that dictate how geometrical and constitutive models influence the complex rheological behavior of the melt are not yet fully understood, which hinders further utilization and optimization. In this study, the three-dimensional (3D) gas–liquid–gas model for catheter gas-assisted extrusion was constructed. Subsequently, the Bird–Carreau model and the Phan–Thien–Tanner (PTT) model were employed in finite element numerical simulations to analyze the complex behavior. For comparative analysis, simplified two-dimensional (2D) model numerical simulations were also conducted. Additionally, experiments on catheter gas-assisted extrusion and parameterization studies of key constitutive model parameters were performed. The findings indicate that the 3D model, when integrated with the PTT constitutive model, demonstrates superior predictability and aligns more closely with experimental results. Furthermore, as the flow rate increases, discrepancies among different models diminish, and the distance required for the melt and gas to achieve motion equilibrium decreases. The internal mechanisms behind these phenomena are elucidated through the analysis of velocity and stress field distributions. This research enhances our understanding of the complex rheological behavior in polymer catheter gas-assisted extrusion, providing valuable insights for both academic research and industrial production in this field.

In the gas-assisted extrusion process, the melt inside the die is in a low-viscosity molten state, so the flow field of the gas cushion layer has a great effect on the cross-sectional shape of the micro-tube. Therefore, this study establishes the gas distribution chamber model of the gas-assisted die. Ansys Fluent software was used to simulate the gas flow field of the gas distribution chamber. The effect of the gas chamber structure on the size of the micro-tube was analyzed by the extrusion experiment. The research shows that the velocity unevenness coefficient of the gas outlet of the single gas chamber die is 11.8%, which is higher than that of the double gas chamber die. The use of a double gas chamber die can improve the stability of the gas cushion layer and the wall thickness non-uniformity of the micro-tube, which verifies the simulation results.