The enhanced infrared shielding function of antimony tin oxide (ATO)-doped polymer-dispersed liquid crystal (PDLC) film enables its use for smart windows, because it can switch between transparent and scattered states, which can protect people’s privacy. When PDLC film is used for a building’s doors and windows or external walls, we hope that it can have a higher infrared shielding capability, in order to reduce the indoor temperature affected by solar irradiation, so as to reduce the energy consumption caused by refrigeration equipment. However, the infrared shielding capability of the existing PDLC is far from sufficient. In this work, modified ATO nanoparticles of different sizes were introduced into the PDLC system to improve its infrared shielding capability. It was found that when the ATO particle size is 20 nm and the doping content is 0.6 wt%, the modified PDLC sample provides optimal infrared shielding function while maintaining excellent electro-optical properties.

Traditional teaching quality evaluation lacks comprehensive study; this paper constructed an evaluation index system of network teaching quality based on multi-source data. Against the background of multi-source data driving, the network teaching quality evaluation index system is constructed, which consists of five indexes: teaching platform function, teaching technology support, teaching service guarantee, teaching process, and teaching feedback. Based on probabilistic hesitation fuzzy sets and evidential reasoning, a network teaching quality evaluation model is established, and probabilistic hesitation fuzzy sets are used to describe the mixed assessment information uniformly. Probabilistic hesitation, fuzzy entropy, and expert weighting are used to assign weights to evaluation indexes. The example verification demonstrated that combining evidential reasoning with utility theory can effectively assess the quality of network teaching. The evaluation results are ranked, and the comparison indicates that this method enhances the accuracy of teaching quality evaluation.

With the aim of effectively formulating recommendations for enhancement and propelling the rapid advancement of these enterprises, this study has formulated a pivotal indicator framework for evaluating innovation capabilities. This framework encompasses six significant categories of indicators, including technological, product, and market innovations, derived from a synthesis of research and case study outcomes. Leveraging the entropy weight method in conjunction with gray correlation analysis, an innovation capacity assessment model was constructed to ascertain indicator weights and their correlation levels in the evaluation process, while holistically considering the influence of diverse factors on enterprise innovation potential. Through empirical analysis involving 20 protective clothing manufacturing enterprises, the outcomes underscored the superior accuracy and reliability of assessment results yielded by the entropy weight-gray correlation model compared to existing methodologies.

To effectively promote police-enterprise cooperation, this study conducted longitudinal multi-case research on existing mechanisms for police equipment technology collaboration through theoretical sampling, with the aim of establishing a process mechanism for equipment technology updates and enhancing the architecture of the police-enterprise cooperation system. The findings highlighted that improving the mechanism for updating police system equipment technology relies on enhancing and establishing mechanisms for demand, management guarantee, cooperation, sharing, and risk prevention and control between police and enterprises. By implementing an update mechanism, it becomes possible to optimize collaboration between the two parties, resulting in enhanced technical sophistication of police equipment.

Smart manufacturing and logistics robots require integrated navigation capabilities. Optimized global routing, fast obstacle avoidance, and stable trajectory generation are all necessary. In this paper, we develop a hybrid path planner. An enhanced A* algorithm is combined with the Dynamic Window Approach (DWA). We intend to satisfy such practical demands. Four key aspects are addressed to optimize the conventional A* algorithm. Evaluation function optimization, neighborhood search optimization, Floyd based path smoothing, and segmented path processing. Path smoothing is also enhanced. These modifications cut computational costs. Invalid node expansion is reduced. Sudden steering changes are mitigated as well. Comparative simulations on two test maps were carried out. The superiority of our improved method was validated. The results show that path turns drop by 33.3% and 37.5%, respectively. Search time is reduced by 22.2% and 31.3%, respectively. Unlike pure static planning strategies, the hybrid framework delivers faster system response. Improved real-time obstacle avoidance capability is also achieved. Consequently, it fits well with practical industrial scenarios. Automated material handling and intelligent warehouse logistics in smart factories are covered.

Addressing the prevalent challenges of inadequate detection accuracy and sluggish detection speed encountered by police patrol robots during security patrols, we propose an innovative algorithm based on the YOLOv8 model. Our method consists of substituting the backbone network of YOLOv8 with FasterNet. As a result, the model’s ability to identify accurately is enhanced, and its computational performance is improved. Additionally, the extraction of geographical data becomes more efficient. In addition, we introduce the BiFormer attention mechanism, incorporating dynamic sparse attention to significantly improve algorithm performance and computational efficiency. Furthermore, to bolster the regression performance of bounding boxes and enhance detection robustness, we utilize Wise-IoU as the loss function. Through experimentation across three perilous police scenarios—fighting, knife threats, and gun incidents—we demonstrate the efficacy of our proposed algorithm. The results indicate notable improvements over the original model, with enhancements of 2.42% and 5.83% in detection accuracy and speed for behavioral recognition of fighting, 2.87% and 4.67% for knife threat detection, and 3.01% and 4.91% for gun-related situation detection, respectively.