Harbin Science and Technology Bureau

The utilization of waste and renewable oils as asphalt modifiers is a crucial strategy for achieving sustainable development in pavement engineering. However, the different physicochemical effects exerted by oil sources (bio-based versus petroleum-based) on the asphalt–aggregate interface remain insufficiently understood. This study aims to elucidate the influence mechanism of two bio-based oils and two petroleum-based oils on asphalt adhesion and the pavement performance of mixtures. A quantitative evaluation method combining the boiling test with digital image processing (DIP) technology was developed to assess the anti-stripping performance of modified asphalt on different lithological aggregates (acidic granite and alkaline limestone). Additionally, Fourier transform infrared spectroscopy (FTIR) was employed to reveal the chemical evolution of the modified asphalt. The results indicated that, although all oil-based modifiers demonstrated excellent compatibility and storage stability with the base asphalt (segregation ratio < 5%), their adhesion properties were significantly influenced by aggregate lithology. The key finding was that, compared to petroleum-based oils, bio-based oils exhibited superior adhesion performance on acidic granite surfaces, markedly mitigating moisture-induced stripping. FTIR analysis confirmed that this enhancement was attributable to the aromatic and carbonyl functional groups introduced by bio-based oils, which effectively promoted the interfacial bonding. Furthermore, bio-oil-modified mixtures exhibited optimal low-temperature cracking resistance without compromising high-temperature stability. These findings elucidate the mechanism by which bio-oil enhances the water-damage resistance of acidic aggregate systems, providing a theoretical basis for the optimized selection of sustainable asphalt modifiers.

Advances in cloud computing and IoT have driven the optimization of distributed systems, enabling near-instantaneous responses. Mobile Edge Computing (MEC) enhances 6 G networks by migrating resource-intensive tasks to edge servers, reducing latency and energy use. This study integrates task offloading and resource allocation to minimize system-wide costs. Although Deep Reinforcement Learning (DRL) is adept at balancing immediate and long-term objectives, resource constraints across mobile devices and servers complicate its application. We propose multi-agent deep reinforcement learning with attention mechanism (AM_MADRL), an attention-based DRL framework, to address these challenges. The algorithm leverages attention mechanisms to extract task features and predict resource demands, enabling MEC servers to make coordinated offloading decisions while respecting system constraints. Simulations confirm AM_MADRL’s superiority over existing methods in cost efficiency and scalability, positioning it as a robust solution for dynamic edge networks.

Simultaneously using infrared and optical images for target detection is becoming a research hotspot. The technology of using multi-modal data for high-performance target detection has proved its great competitiveness in practical applications. However, infrared images and optical images have different emphasis on dominant information, and direct fusion or symmetric feature processing may reduce the learning efficiency of the network, thus affecting the detection performance. To solve these problems, a new fusion detection method based on modal-specific information propagation (MSIP) is proposed in this paper. It mainly includes two innovations: a spatial information propagation module (SaIDM) is designed to efficiently propagate a large amount of spatial information in optical images to the infrared branch. At the same time, a semantic information propagation module (SeIDM) is designed, which uses the semantic information in the infrared branch to enhance the attention of the optical branch. The experimental results on the DroneVehicle dataset show that the network can improve the detection accuracy of difficult targets by transmitting modal dominance information, and the mAP reaches 71.13, which is in the leading level.

How to optimize multi-objective drainage efficiently and accurately is a key problem in urban drainage. In response to this problem, in this paper, a novel method combining geographic information System (GIS), Random Forest Regression (RFR) and Multi-Objective Optimization is proposed. Multi-Objective Forest Regression (MOFR) method provides detailed spatial data and terrain information through GIS, and RFR realizes high-precision prediction of urban drainage system performance, and uses MOO for multi-objective optimization, so as to ensure high prediction accuracy and improve the comprehensive performance of the system. Ultimately, experimental results, obtained through comparative experiments and analysis with traditional machine learning methods, demonstrate that the MOFR method exhibits notable advantages in enhancing prediction accuracy, minimizing function errors, and optimizing water quantity testing. Especially in complex multi-objective optimization scenarios, MOFR method can more accurately find the optimal design scheme and reduce the uncertainty in the decision-making process. This fully verifies the reliability and effectiveness of the architecture in urban drainage system optimization design.

BACKGROUNDIntracranial epidermoid cyst (IEC) transformation to malignant squamous cell carcinoma (SCC) is extremely rare, and its etiology is yet unknown. Currently, SCC is treated by performing surgery, followed by a combination of radiotherapy and chemotherapy. It is crucial to identify efficient and trustworthy therapeutic targets for SCC to improve its diagnosis, prognosis, and treatment.CASE SUMMARYIn this study, we report the case of a 47-year-old female patient with SCC, which progressed from IEC in the left internal capsule region. The patient was sought treatment at our hospital for severe diplopic vision, accompanied with speech disorder and memory loss. Based on the clinical and postoperative pathology, this patient was finally diagnosed with SCC. To identify disease-causing variants, whole exome sequencing (WES) was performed on the proband. WES revealed two pathogenic missense mutations on Gap junction protein beta 2 (GJB2) (c.257C>T) and Toll-like receptor 2 (TLR2) (c.1039A>G), respectively.CONCLUSIONThis study provided the first clinical evidence for demonstrating the role of GJB2 and TLR2 in IEC development and treatment. We further confirmed WES as a robust and reliable technique for underlying rare and complex disease-related genetic factor identification.

Change detection (CD) is a crucial application in the field of remote sensing. Most current CD methods are based on deep learning and revolve around multispectral data. However, a common issue arising from these methods is the large number of model parameters due to the high dimensionality of the data channels. In this paper, we propose a lightweight Siamese network structure that minimally incorporates conventional convolutional layers. Additionally, in terms of data preprocessing, we select the R, G, and B channels of multispectral data for dehazing processing, and employ the processed data as inputs to the network. Experimental results illustrate the outstanding effectiveness and efficiency of the proposed method.

A comprehensive understanding of the evolution of soybean climate potential productivity and its response to climate change in Heilongjiang Province can offer reference and basis for further tapping soybean production potential and realizing stable and high yield of soybean in the frigid region. Based on meteorological data from 80 meteorological stations in Heilongjiang Province from 1961 to 2020, we estimated photosynthesis, light temperature, and climate potential productivity of soybean by the stepwise correction method, examined the spatiotemporal variations by spatial interpolation and statistical analysis methods, and analyzed the impact of changes in climate factors such as radiation, temperature, and precipitation on climate potential productivity. The results showed that during the study period, the average values of photosynthesis potential productivity (YQ), light-temperature potential productivity (YT), and climate potential productivity (YW) of soybean in Heilongjiang Province were 7533, 6444, and 3515 kg·hm-2, respectively. The temporal changes of those variables showed significant increasing trends, with increases of 125.9, 182.9, and 116.1 kg·hm-2·(10 a)-1, respectively. For the spatial distribution, YQ, YT, YW were characterized by high values in plains and lower in the mountains, and gradually decreased from southwest to northeast. Compared with that during 1961-1990, the high value zone of YW in period 1991-2020 expanded by 7.1%, and the low value zone decreased by 5.1%. YW showed a significant response to climate change. The potential temperature growth period was extended due to climate warming. The continuous increase in thermal resources, combined with relatively sufficient precipitation, effectively alleviated the negative impact of the decline in light resources on soybean production in Heilongjiang Province. The projected "warm and humid" climate would comprehensively boost climate potential productivity of soybean in Heilongjiang Province.