Search

In dynamic environments, achieving accurate and robust Visual SLAM (Simultaneous Localization and Mapping) remains a significant challenge, particularly for applications in robotic navigation and autonomous driving. This study introduces YG-SLAM, an innovative approach that integrates YOLOv8 and geometric constraints within the ORB-SLAM2 framework to adapt effectively to dynamic scenarios.YOLOv8 is employed for instance segmentation and dynamic object detection, enriching the semantic information while extracting image feature points. Geometric constraints, including epipolar geometry algorithms and Lucas-Kanade optical flow methods, are utilized to filter out dynamic objects effectively.The tracking thread exclusively relies on static feature points for camera pose estimation, substantially improving the system’s localization accuracy. Experimental results on the TUM dataset demonstrate that YG-SLAM significantly outperforms traditional ORB-SLAM2 in dynamic environments. Specifically, the Root Mean Square Error (RMSE) of the absolute trajectory errors reduced by 96.51% in comparison to ORB-SLAM2, and the RMSE of the relative pose errors decreased by 93.60% when compared to the performance of ORB-SLAM2.These notable reductions in errors demonstrate the promising performance enhancements of YG-SLAM over traditional ORB-SLAM2 in dynamic environments.

The bandwidth and sensitivity of traditional terahertz detection devices are limited by material properties and working temperature. Here, we proposed a four-level Rydberg atom detection method, which can realize real-time detection for the absolute strength of the terahertz electric field under room temperature with the detection bandwidth of 0-5 THz. By theoretical analysis and formula derivation, the relation among the terahertz electric field strength, excitation energy level, coupling photoelectric field strength, gas cell temperature and electromagnetically induced transparency (EIT) signal are studied in detailed. The NEP of this method is proven to be 10−17 W/Hz1/2 in the range of 0-1 THz, which is three orders of magnitude smaller than that of the traditional optimal thermal detection system. This research has important reference value for the later high-sensitivity detection and application development of terahertz waves

Single-photon avalanche diodes (SPADs) are widely used for weak light detection due to their high gain and high sensitivity. Unfortunately, SPAD devices have random telegraph signal (RTS) noise during the avalanche transition phase, which makes circuit design more difficult. Therefore, the modeling of RTS noise in SPAD devices is of great significance for the design of signal processing circuits. However, RTS noise is not included in the traditional equivalent circuit model. In response to the above problems, this paper implements an equivalent circuit model with RTS noise based on the Cadence. The model is built based on the basic components in the library, and it has strong application universality. The model is validated in a novel embedded photogate SPAD device (EP-SPAD). It accurately simulates the RTS noise and IV characteristics of the EP-SPAD device. In addition, the threshold voltage of EP-SPAD device operating in Geiger mode is precisely defined based on the characteristics of RTS noise. When the reverse bias voltage is increased from 0 V to 15 V, the IV characteristics of the model are consistent with the EP-SPAD device.

The location of logistics nodes that facilitate the China Railway Express (CRexpress) and rail-sea intermodal transportation has received increasing attention in China under the Belt and Road initiative. This paper selects 38 cities as candidate Chinese international container intermodal hubs (CICIHs) through qualitative screening by considering the current operations of the CRexpress/rail-sea intermodal transportation and government planning strategies. Subsequently, five connectivity indexes are proposed to reflect the node performance considering the multiple modes and stages of transportation in a logistics chain. In view of the shortcomings of the grey relational analysis-technique for order preference by similarity to ideal solution (GRA-TOPSIS) method, a hybrid multi-criteria decision making (MCDM) model is proposed to comprehensively evaluate the hub location. The model is based on the grey area relational analysis-technique for order preference by similarity to ideal solution (GARA-TOPSIS) in combination with analytic hierarchy process, entropy method and game theory. Tianjin, Harbin, Zhengzhou, Wuhan, Chongqing, Xi’an, Urumqi, Guangzhou, Shanghai and Nanning are selected as the final CICIHs, considering the geographical location, economic development and national policy pertaining to the Belt and Road initiative. A sensitivity analysis of the index weight and preference coefficient and a comparative analysis of different MCDM methods are performed. The robustness and stability of the proposed model are verified. This study can support the location selection of CICIHs and expand the methods and applications in the decision-making field.

The COVID-19 global pandemic has posed a serious impact on human life and health. Facing this unknown, unexpected, and devastating disease, China launched a resolute battle to prevent and control its spread. This study focused on assessing the recovery of city activity of 17 administrative regions in China around the COVID-19 pandemic period in early 2020. The night-time light (NTL) satellite remote sensing data, NPP-VIIRS dataset, was applied to evaluate the NTL changes during the coronavirus epidemic to understand the impact of the epidemic on city activity and the economy of China. A series of median composites was first constructed to analyze short-term NTL changes. Based on these median composites, the ratio map, difference map, wave index, and recovery ratio (RR) were defined to indicate the NTL recovery pattern and its spatial distribution characteristics. The results are as follows:first, the adopted median composites are able to describe the NTL changes in the short term; second, the recovery of city activity could not be exactly described by RR of a single year but could be evaluated by the comparison of RR between two adjacent years; third, a partial and spatially heterogeneous recovery of city activity is observed after the 2020 holiday compared to 2019, fourth, a better degree of recovery is observed after China announced its fully work resumption.

Caffeine is an alkaloid and may be the most commonly ingested pharmacologically active substance in the world, but continuous abuse may lead to “caffeine poisoning”. In this study, we propose a method for the qualitative and quantitative analysis of caffeine in various medicines using terahertz spectroscopy combined with chemometrics. By comparing this terahertz (THz) spectroscopy technology (Fourier transform infrared instrument, FTIR) with high performance liquid chromatography (HPLC) and Raman spectroscopy, we prove that there is less than a 5% difference between the THz and HPLC results, which is far superior to the nondestructive testing results obtained using Raman spectroscopy. In addition, the quantitative analysis of caffeine in 86 medicines was conducted using the support vector regression (SVR) chemometric method, and the correlation coefficient R achieves 99.61%. Therefore, we have effectively proven that THz spectroscopy technology combined with chemometric can achieve nondestructive, fast, and efficient qualitative and quantitative detection of key ingredients in medicines.

The existing sequential recommendation algorithms cannot effectively capture and solve the problems such as the dynamic preferences of users over time. This paper proposes a deep Recommendation model CLSR (Combines Long-term and Short-term interest Recommendation) that Combines long-term and short-term interest preferences. Firstly, the model models the potential feature representation of users and items, and uses the self-attention mechanism to capture the relationship between items in the interaction of users’ historical behavior, so as to better learn the short-term interest representation of users. At the same time, the BiGRU network is used to extract the features of users’ long-term interests on a deep level. Finally, the features of long-term and short-term interest are fused. On four publicly available datasets, experimental results show that the proposed method has better improvement on HR@N, NDCG@N and MRR@N, which validates the effectiveness of the model.

Single photon avalanche diodes (SPAD) based on avalanche effect have been widely used in the detection of extremely weak light signals. Conventional SPAD devices manufactured with silicon-based CMOS technology have a high dark count rate (DCR), making it difficult to accurately detect single photon signals. This paper proposes a low dark count rate ring-gate SPAD (RG-SPAD) to solve above problems, and compares RG-SPAD with conventional SPAD (C-SPAD) and conventional dummy-gate SPAD (CDG-SPAD). Slivaco-Technology Computer Aided Design (TCAD) performs two-dimensional simulation of the device to verify the basic principles of SPAD. Three types of SPAD device are manufactured by standard BCD process. The passive quenching circuit is built to obtain the DCR of SPAD devices. The avalanche breakdown voltages of C-SPAD, CDG-SPAD and RG-SPAD are 11.55 V, 11.85 V and 11.4 V respectively. In order to test the dark count rate of SPAD devices, an over-bias voltage of 1 V was applied to three types of SPAD device under the same size condition (temperature: 22 °C). The DCR of RG-SPAD (186 Hz) with ring-gate structure is significantly lower than CDG-SPAD (378 Hz) with conventional dummy-gate structure and C-SPAD (498 Hz) with conventional structure. When the room temperature dropped to 18 °C, RG-SPAD still maintained the lowest DCR (148 Hz) among three types of device structure.

This paper proposes a robust dual-color watermarking based on quaternion singular value decomposition (QSVD), which can embed large payloads into color images with low distortion, and can obtain strong robustness to process color image in a holistic manner. First, two notes are proposed for designing the proposed watermarking scheme, one of which is about three strong correlations found in U that can be used for watermark embedding, and the other is analyzing the feasibility of V compensation for QSVD-based watermarking scheme. In addition, a fast structure-preserving algorithm is used to calculate the singular value decomposition (SVD) of a quaternion matrix, which makes the procedure computationally more flexible and efficient. Then a new watermarking scheme is proposed to protect the copyright of color images. This scheme uses quaternion to make the color image channels correlated so that the proposed watermarking scheme has strong anti-attack performance. Experimental results show that the proposed dual-color watermarking is not only imperceptible but also robust to some common attacks, and the performance of the proposed method outperforms other methods considered in this work.

Recent studies indicated that numerous long noncoding RNAs (lncRNAs) are closely related to human diseases and can serve as potential biomarkers and drug targets for complex diseases. Therefore, identifying lncRNAs associated with diseases through computational methods is conducive to the exploration of disease pathogenesis. Most previous studies had shortcomings, such as low prediction accuracy, the need for negative samples, and weak generalization. Such studies established shallow prediction models and failed to fully capture the complex relationships among lncRNA-disease associations, lncRNA similarity, and disease similarity. LRLSSP, a new computational method based on Laplacian regularized least squares (LRLS) and space projection was used to predict candidate disease lncRNAs in this study. LRLSSP deeply integrates information on lncRNA similarity, disease similarity, and known lncRNA-disease associations. The estimated score of lncRNA-disease association was obtained through LRLS, and network projection was utilized to reliably predict disease-related lncRNAs. Leave-one-out cross validation(LOOCV) was implemented to evaluate the prediction performance of LRLSSP. Results showed that LRLSSP performed was better than other state-of-the-art methods in predicting lncRNA-disease associations. In addition, case studies conducted on melanoma,cervical cancer, ovarian cancer and breast cancer indicated that LRLSSP can discover potential and novel lncRNA-disease associations. Overall, the results demonstrated that LRLSSP may serve as a reliable and effective computational tool for disease-related lncRNAs prediction.

With the integration of large-scale renewable energy and various new types of loads, the stochastic fluctuation characteristic of the power has seriously affected the secure and stable operation of interconnected power grids. In addition to active power fluctuations of AC tie-lines that have been studied in the past, the problem of dynamic fluctuation of bus voltage is also increasingly prominent. Firstly, the typical power fluctuations of electric railway traction load, smelting load, and wind power generation in modern power systems are presented and the accompanying voltage fluctuations are analyzed. Secondly, the dynamic fluctuation mechanism and distribution characteristic of the system voltage under special forced oscillation and generalized forced oscillation are studied. It is found that when the periodic disturbance induces a special forced oscillation, the voltage fluctuation has the same mode information compared to the power angle fluctuation. And it also has obvious peak characteristics near the natural frequency. When the stochastic disturbance induces generalized forced oscillation, the voltage fluctuations can be decomposed into proportional and resonant components according to its frequency domain characteristic. Finally, the above conclusions are verified by the simulation results of an IEEE 4-generator 2-area system.

As semiconductor process continues to advance, the miniaturization of feature sizes places higher demands on high-failure electro-static discharge (ESD) applications. This article explores the connection between the physical structure of a device-level silicon controlled rectifier (SCR) and high-failure ESD characteristics. The gate-controlled silicon controlled rectifier (GCSCR) based on the gate control effect is fabricated using the $0.18~\mu \text{m}$ standard bipolar complementary-metal-oxide-semiconductor double-diffused-metal-oxide-semiconductor (BCD) process. The ESD characteristics of the device are analyzed by technology computer aided design (TCAD) simulation and equivalent circuits. The transmission line pulse (TLP) is used to test the performance of the device. The results show that when the gate length is $4~\mu \text{m}$ , the failure current of the device is only 1.56A. When the gate length is $1~\mu \text{m}$ , the trigger voltage and the holding voltage of the device are 24.4V and 21.1V respectively, and the failure current is 34.94A. According to the test results of the above devices, it can be concluded that the current release mode of GCSCR with different gate sizes significantly affects the ESD characteristics of the device.

In this work, we focused on the resonant cavity effect of the air plasma, formed during femtosecond laser filamentation, on the near- and far-field profiles of the generated terahertz (THz) radiation. Via the method of ray tracing, it has been demonstrated that both on-axis propagation and conical forward emission of THz wave could be qualitatively interpreted in case of considering the filament as a Fabry-Perot-like micro-cavity with radial step refractive index distribution in THz domain. Our theory might potentially solve the contradictions of different experimental observations on THz spatial profiles reported in publications, and renew the understanding of THz wave propagation inside the filament plasma column.