Compared with other materials, polyethylene terephthalate (PET) has high transparency, excellent physical and mechanical properties in a wide temperature range and good hygiene and safety, so it is widely used in the packaging industry, especially in the packaging of beverages and foods. The optimization of PET bottles is mainly reflected in three aspects: material optimization, structure optimization and process optimization, among which there is much research on material optimization and process optimization, but there is no complete overview on structure optimization. A summary of structural optimization is necessary. Aiming at structural optimization, the finite element method is a useful supplement to the beverage packaging industry. By combining the computer-aided design technology and using finite element software for finite element simulation, researchers can replace the experimental test in the pre-research design stage, predict the effect and save cost. This review summarizes the development of PET bottles for beverage packaging, summarizes various optimization methods for preventing stress cracking in beverage packaging, and especially focuses on comparing and evaluating the effects of several optimization methods for packaging structure. Finally, the future development of all kinds of optimization based on structural optimization in the field of beverage packaging is comprehensively discussed, including personalized design, the combination of various methods and the introduction of actual impact factor calculation.

PET bottlesare often used as airtight containers for filling carbonated drinks. Because carbonated drinks contain large volumes of CO2 gas, the container needs to bear a tremendous pressure from the inside of the bottle.If the stress exceeds the bearing limit, the material will show the phenomenon of local cracking and liquid overflow.For the structural design, the method of manual adjustment before automatic adjustment was adopted. First, through manual optimization, the initial optimal parameter combination was as follows:the inner diameter of the bottle bottom was 17 mm, the dip angle of the valley bottom was 81°, the deepest part of the valley bottom was 25 mm, and the outer diameter was 27 mm. Comsol software was used for automatic optimization. Compared with the original bottle bottom, the total maximum principal stress and total elastic strain energy in the bottle bottom after manual–automatic double optimization decreased by 69.4% and 40.0%, respectively, and the displacement caused by deformation decreased by 0.60 mm (74.1%). The extremely high reduction ratio was caused by manual–automatic double optimization.

Accurate monitoring of fire and smoke plays an irreplaceable role in preventing fires and safeguarding the safety of citizens' lives and property. The network structure of YOLOv5 is simple, but using convolution to extract features will lead to some problems such as limited receptive field, poor feature extraction ability, and insufficient feature integration. In view of the current defects of YOLOv5 target detection algorithm, a new algorithm model named Swin-YOLOv5 was proposed in this work. Swin transformation mechanism was introduced into YOLOv5 network, which enhanced the receptive field and feature extraction ability of the model without changing the depth of the model. In order to enrich the feature map splicing method of weighted Concat and enhance the feature fusion ability of model pairs, the feature splicing method of three output heads of feature fusion layer network was improved. The feature fusion module was further modified, and the weighted feature splicing method was introduced to improve the network feature fusion ability. Experimental results show that the map (average rage accuracy) of this method rises faster than the benchmark algorithm. Under the same experimental dataset, the map of this algorithm is improved by 0.7%, and the high-precision target detection speed is improved by 1.8 FPS (fast packet switch). Under the same experimental dataset, the improved algorithm could more accurately detect the targets that were not detected or detected inaccurately by the original algorithm, which embodied the adaptability of real scene detection and had practical significance. This work provided an opportunity for the application of fire-smoke detection in forest and indoor scenes and also developed a feasible idea for feature extraction and fusion of YOLOv5.