Hechi University

Wang B., Wang R., Chen Y., Yang C., Teng X., Sun P.

Wood surface defect detection is a critical step in wood processing and manufacturing. To address the performance degradation caused by small targets and multi-scale features in wood surface defect detection, a novel deep learning model is proposed in this study, FDD-YOLO, specifically designed for this task. In the feature extraction stage, the C2f module and the funnel attention (FA) mechanisms are integrated into the design of the C2f-FA module to enhance the model’s ability to extract features of wood surface defects of various sizes. Additionally, the Dual Spatial Pyramid Pooling-Fast (DSPPF) module is developed, and the Context Self-attention Module (CSAM) is introduced to address the limitations of traditional max pooling methods, which often overlook global contextual information when extracting local features, thereby improving the detection of small-scale wood defects. In the feature fusion stage, a Dual Cross-scale Weighted Feature-fusion (DCWF) module is proposed to fuse shallow, deep, and cross-scale features through a weighted summation approach, effectively addressing the challenge of scale variation in wood surface defects. Experimental results demonstrate that the proposed FDD-YOLO model significantly improves detection performance, increasing the mAP of the baseline model YOLOv8 from 78% to 82.3%, a substantial enhancement of 4.3 percentage points. Furthermore, FDD-YOLO outperforms other mainstream defect detection models in terms of detection accuracy. The proposed model demonstrates significant potential for industrial applications by improving detection accuracy, enhancing production efficiency, and reducing material waste, thereby advancing quality control in wood processing and manufacturing.

Wang C., Zhou Z., Peng J., Dai X.

The consensus problem under a single leader forms the foundation of coordination control in multi-agent systems (MASs). Nonetheless, the issue of containment control, which involves multiple leaders, represents a critical aspect that demands further investigation. This paper addresses the containment control of nonlinear heterogeneous partial difference multi-agent systems (HPDMASs) characterized by state time delays and governed by hyperbolic or parabolic partial difference equations. This study proposes a D-type distributed iterative containment control protocol incorporating initial state learning, which leverages local state information from neighboring agents. Using the contraction mapping principle, this research examines the convergence conditions for the containment error within a finite time interval. The results demonstrate that, under the proposed containment control protocol, the containment error in the HPDMASs can gradually converge to zero as the number of iterations increases. Simulation results are provided to validate the effectiveness of the proposed control protocol.

Sun H., Zhang X., Liu X., Xiong S.

The estimation of the back electromotive force (EMF) signal of the permanent magnet synchronous motor (PMSM) using sliding mode observer (SMO) has a large number of high-frequency chattering, which cannot be directly applied to the sensorless control system of PMSM. This article proposes a robust sensorless control scheme based on an improved back EMF observer and an extended state observer (ESO). First, a super-twisting algorithm back EMF observer (STABEO) is proposed by combining the back EMF model and the super-twisting algorithm, in which the equivalent signal of the back EMF is estimated through the traditional SMO. The STABEO takes into account the uncertainty and unmodeled dynamics in the back EMF model, which leads to smaller chattering, better estimation and faster convergence. Then, an ESO is constructed to calculate the rotor position, the rotor speed and the total interference, and after that, a robust speed controller is designed. In addition, the stability is proved through the Lyapunov stability theory. Finally, the effectiveness and feasibility of the proposed control scheme were verified through experiments. The proposed robust sensorless control scheme has the characteristics of strong anti-interference ability, faster convergence, smaller high-frequency chattering and without losing its robustness.

Zhang K., Song J., Lu J., Zhao L., Deng W., Guan D., Mao B.

Yunnan Province, a global biodiversity hotspot, hosts a diverse array of Acrididae grasshoppers essential for ecosystem dynamics and agriculture. To elucidate the phylogenetic relationships and evolutionary history of this group, we sequenced and analyzed complete mitochondrial genomes from 17 Acrididae species endemic to Yunnan, with genome lengths ranging from 15,403 to 15,943 base pairs. These data were integrated with mitochondrial sequences from 46 additional species to construct comprehensive phylogenetic trees. The maximum-likelihood tree identified four major clades with robust support (bootstrap values > 90%), revealing significant lineage diversification during the Early Eocene (51.94 million years ago, Mya) and subsequent radiations in the Miocene (~20 Mya) and Pliocene (~5 Mya). These divergence times correlate with major geological events and climatic shifts in the region, such as the uplift of the Tibetan Plateau and the intensification of the Asian monsoon. Notably, several species within the Coptacrinae and Oxyinae subfamilies, particularly Pseudodotraulia cornuata and Spathosternum prasiniferum, were found to be polyphyletic, indicating the necessity for taxonomic revisions. Further cluster analyses of codon usage bias and genetic distance support these taxonomic revisions within Acrididae. This robust phylogenetic framework underpins conservation strategies aimed at preserving Yunnan’s rich grasshopper biodiversity and informs updates to their phylogeny.

Guan D., Qin Y., Chen Y., Zhang S., Liu J., Yi H., Li X.

The mulberry looper (Phthonandria atrilineata), a geometrid moth, plays a pivotal role in the destruction of mulberry trees (Morus spp.). In China, P. atrilineata is the most significant insect pest to sericulture, as it feeds on mulberry leaves and spreads diseases. The outbreak trend of P. atrilineata has been expanding yearly, causing substantial economic losses. Despite its ecological and economic importance, knowledge about the genomic background of P. atrilineata remains limited. Here, we report a chromosome-level reference genome of P. atrilineata, with a total size of 336.55 Mb, containing 15,026 protein-coding genes and 39.72% repeat sequences. These findings have the potential to shed light on the genetic basis of the destructive nature and environmental adaptation of P. atrilineata, offering valuable genomic resources for understanding genome evolution and pest management within this Lepidopteran pest.

Yang L., Wu J., Luo L., Wu W., Ma H.

Wu H., Huang J., Qin Y., Sun Y.

This paper investigates the state estimation problem for nonlinear cyber–physical systems (CPSs). To conserve system resources, we propose a novel hybrid dynamic event-triggered mechanism (ETM) that prevents the occurrence of Zeno behavior. This work is based on designing an interval observer under the hybrid dynamic ETM to solve the state reconstruction problem of Lipschitz nonlinear CPSs subject to disturbances. That is, the designed triggering mechanism is integrated into the design of the Interval Observer (IO), resulting in a hybrid dynamic event-triggered interval observer (HDETIO), and the system stability and robustness are proved using a Lyapunov function, demonstrating that the observer can effectively provide interval estimation for CPSs with nonlinearity and disturbances. Compared to existing work, the primary contribution of this work is its ability to pre-specify the minimum inter-event time (MIET) and apply it to interval state estimation, enhancing its practicality for real-world physical systems. Finally, the correctness and effectiveness of the designed hybrid dynamic ETM and IO framework are validated with an example.

Zhang X., Li Y., Xiong S., Liu X., Guo R.

This study addresses the completely distributed consensus control problem for the heterogeneous nonlinear multi-agent system (MAS) with disturbances under switching topology. First, a global sliding mode manifold (GSMM) is designed for the overall MAS dynamic, which maintains stability without oscillations during topology switching after achieving the sliding mode. Subsequently, a consensus sliding mode control protocol (SMCP) is proposed, adopting the common sliding mode control (SMC) format and ensuring the finite-time reachability of the GSMM under topology switching. Finally, the proposed GSMM and SMCP are applied to the formation control of multiple-wheeled mobile robots (WMRs), and simulation results confirm their feasibility and effectiveness. The proposed SMCP design demonstrates key advantages, including a simple control structure, complete robustness to matched disturbance, and reduced-order dynamics under the sliding mode.

Feng X., Huang Z., Ahmad S., Ma C., Zhu J., Li K., Peng G., Tettamanti G., Hu W., Xu H., Tian L.

Abstract The black soldier fly larvae (BSFL) have garnered great attention for their potential role in converting organic waste into high-quality insect proteins, providing valuable feed components for animal production and highly enriched organic manure for crop production. However, environmental factors such as high temperatures can compromise their productivity. To address this, we conducted selective breeding of BSFL at 40 °C, and the bred group was improved with higher body weight, survival rates, and substrate reduction after nine generations of breeding. Notably, the gut microbiota of the heat-tolerant colony was highly represented by the bacteria belonging to Proteobacteria, Firmicutes, and Actinobacteria, while the Bacteroidetes phyla were decreased. Transcriptomic analysis revealed that the upregulated genes in 40 °C bred groups were primarily involved in biological processes such as Spliceosome, FoxO signaling pathway, longevity regulating pathways, Drug metabolism, and Xenobiotic metabolism via cytochrome P450. These findings provide a detailed understanding of the molecular mechanisms underlying BSFL’s adaptation to heat stress, facilitating future industrial application advancements.

Zhou J., Zeng J., Zhou L., Jiao H., Huang Y., Ouyang L., Zhu R., Long Y.

The dynamic evolution of cavitation bubbles in underwater wet laser welding significantly influences the laser beam. To mitigate the adverse effects of cavitation bubbles, an ultrasonic-assisted underwater laser wet welding method is proposed. This work reports an experimental study of ultrasonic-assisted underwater laser welding 304 stainless steel at varying welding speeds, different water layer thicknesses, and different ultrasonic powers. Additionally, the high-speed photography method was used to analyze the effect of ultrasound on the movement of cavitation bubbles, and a ray optics finite element model was employed to better understand the impact of cavitation positions on the laser beam. The results indicate that, under identical process parameters, the application of ultrasonic waves controls the movement of cavitation bubbles via acoustic flow. Consequently, the weld seam is more uniform, and the width and depth are increased effectively. In addition, without ultrasonic assistance, the weld seam width and depth decrease with higher welding speeds and decrease with the increase of the water layer thickness. With ultrasonic assistance, the weld seam width decreases while the depth increases at higher welding speeds. The weld seam width and depth increase with the increase of ultrasonic power. The weld width first decreases and then increases as the water layer thickness increases, and the weld seam depth shows a decreasing trend. These indicate that the dynamic evolution of cavitation bubbles has an important impact on the weld width and depth. This study can provide a reference for achieving better weld efficiency and geometry in underwater laser wet welding.

Zhang G., Zhang W., Ren C., Li J., Lan W., Lai H.

Yi Y., Wang Z., Shi Y., Song Z., Zhao B.

Wei L., Li C., Wen C., Lai H.

2-Phenylquinoline-4-carboxylic acid-Ir(iii) complex (Irphen) was successfully assembled into a Zr-MOF, which showed high efficiency in catalyzing sulfide photooxidation with blue light irradiation under aqueous condition.

Dai L., Zhang G., Ou Y., Liu X., Yao H., Hu W., Li H., Lan W.

One new gliotoxin derivative fumianthrogliotoxin (1), one new indoquizoline alkaloid N3-(methyl propionate) indoquizoline (2), and three novel indole alkaloids, anthroxyindole (3), (±)-asperfumiindole A (4), and (±)-asperfumiindole B (5), together with 16 known compounds (6–21), were isolated from the culture of deep-sea derived fungus Aspergillus fumigatus AF1. Their chemical structures and absolute configurations were determined through the analysis of NMR data in combination with electronic circular dichroism (ECD) calculations and other spectroscopic analyses. Compounds 2–11 and 13–21 were evaluated for anti-pulmonary fibrosis activity. Compounds 8 and 13 displayed significant downregulation of the mRNA expression levels of all three molecular markers (COL1A1, α-SMA and FN1), with compound 13 exhibiting the best performance among all the tested compounds.

Yang H., Su X., Li B., Xia C., Zheng H., Yang M., Zhang T.

Efficient and intelligent real-time roughness detection is essential for industrialized production of high-precision machining and manufacturing industries. The traditional optical inspection methods are designed for offline measurement which is not appropriate for online quick detection. The development of artificial intelligence and machine vision makes online roughness inspection possible. At present most methods utilize deep learning classification tasks, which can only provide one detection result under one field of view and cannot adapt to the inhomogeneous distribution of roughness. To address this issue, we propose a novel semantic segmentation approach processing with white light interference images to segment different roughness values under one field of view at the pixel level, thereby improving the detection results. The strategy and detection mechanism are expressed firstly. The strongest light intensity (0th level) interference fringe is the focus to be segmented. Based on this analysis, a semantic segmentation model, FAMNet, is proposed in this paper for improving surface roughness detection precision. White light interference image is as the input of model. Fusion Attention Mechanism (FAM) algorithm is specifically designed to be embedded into backbone to construct network structure. The channel attention and positional attention are combined together to achieve a balance between detection accuracy and computation time. The channel attention improves the ability of the attention network to perceive contextual information and enhances the connectivity between different channels. The positional attention mechanism reduces the loss of positional information and enhances the scope of utilization of network features. Comparison experiments are executed and the experiment results show that the fusion attention mechanism module algorithm has a better detection ability for white light interference fringe in the balance of accuracy and computing time. The segmentation recognition accuracy MIoU reaches 86.3%, and it improves the performance of computation by about 77.66% compared to the current best-performing model and the amount of parameter is lower than it by about 63.59%. The proposed FAMNet model has a good performance for surface roughness online detection.