基于改进YOLOv8的轻量化小麦病害检测方法

为提高小麦病害检测精度,实现将模型方便快速部署到移动端,该研究提出了一种基于改进YOLOv8的轻量化小麦病害检测方法。首先,使用PP-LCNet模型替换YOLOv8网络结构的骨干网络,并在骨干网络层引入深度可分离卷积(depthwise separable convolution, DepthSepConv)结构,减少模型参数量,提升模型检测性能;其次,在颈部网络部分添加全局注意力机制(global attention mechanism, GAM)模块,强化特征中语义信息和位置信息,提高模型特征融合能力;然后,引入轻量级通用上采样内容感知重组(content-aware reassembly of features,CARAFE)模块,提高模型对重要特征的提取能力;最后,使用Wise-IoU(weighted interpolation of sequential evidence for intersection over union)边界损失函数代替原损失函数,提升网络边界框回归性能和对小目标病害的检测效果。试验结果表明,对于大田环境下所采集的小麦病害数据集,改进后模型的参数量及模型大小相比原YOLOv8n基线模型分别降低了12.5%和11.3%,同时精确度(precision)及平均精度均值(mean average precision,m AP)相较于原模型分别提高了4.5和1.9个百分点,优于其他对比目标检测算法,可为小麦病害检测无人机等移动端检测装备的部署和应用提供参考。

Data-driven methods for predicting the representative temperature of bridge cable based on limited measured data

Cable-stayed bridges have been widely used in high-speed railway infrastructure.The accurate determination of cable’s representative temperatures is vital during the intricate processes of design,construction,and maintenance of cable-stayed bridges.However,the representative temperatures of stayed cables are not specified in the existing design codes.To address this issue,this study investigates the distribution of the cable temperature and determinates its representative temperature.First,an experimental investigation,spanning over a period of one year,was carried out near the bridge site to obtain the temperature data.According to the statistical analysis of the measured data,it reveals that the temperature distribution is generally uniform along the cable cross-section without significant temperature gradient.Then,based on the limited data,the Monte Carlo,the gradient boosted regression trees(GBRT),and univariate linear regression(ULR)methods are employed to predict the cable’s representative temperature throughout the service life.These methods effectively overcome the limitations of insufficient monitoring data and accurately predict the representative temperature of the cables.However,each method has its own advantages and limitations in terms of applicability and accuracy.A comprehensive evaluation of the performance of these methods is conducted,and practical recommendations are provided for their application.The proposed methods and representative temperatures provide a good basis for the operation and maintenance of in-service long-span cable-stayed bridges.

TWO IMPROVED GRAPH-THEORETICAL CLUSTERING ALGORITHMS

Graph-theoretical approaches have been widely used for data clustering and image segmentation recently. The goal of data clustering is to discover the underlying distribution and structural information of the given data, while image segmentation is to partition an image into several non-overlapping regions. Therefore, two popular graph-theoretical clustering methods are analyzed, including the directed tree based data clustering and the minimum spanning tree based image segmentation. There are two contributions： （1） To improve the directed tree based data clustering for image segmentation, （2） To improve the minimum spanning tree based image segmentation for data clustering. The extensive experiments using artificial and real-world data indicate that the improved directed tree based image segmentation can partition images well by preserving enough details, and the improved minimum spanning tree based data clustering can well cluster data in manifold structure.

Security Assurance for One-stage Resection of Left Colon Cancer with Acute Obstruction—Thorough and Prompt Enteral Decompression without Contamination

Objective: Thorough, prompt enteral decompression technique without contamination was de- veloped to ensure safety for emergent colon resection and primary anastomosis. Methods: After isolating the mesentery, the “to be resected colon segment” was cut at its lower end, then the proximal cut end was put into a plastic bag which was adhered to one side of the operating table. After releasing the clamp, the content could ?ow into this bag. The operator could squeeze the bowel with two hands by turns, from proximal to farness, and from small bowel to large bowel, until the entire bowel content was fully discharged. Then the upper end of this “to be resected colon segment” was cut, and was removed together with the plastic bag. Results: 31 cases of left colon cancer with acute obstruction were decompressed with this technique. They all recovered smoothly, without anastomosis ?stula. Another 6 cases of hepatic seg- mentectomy with incidental colonectomy were decompressed with this technique and had the same results. This technique was also used in di?erent kinds of acute small intestinal obstruction and gained satisfactory results. Conclusion: This technique could be considered as the preferable choice for intraoperative enteral decompression.

Determination of energy coupling to material in laser welding by a novel “sandwich” method

A mathematical energy coupling model was developed to analyze the light transmission in the keyhole and energy distribution on the keyhole wall.The main characteristics of the model include：1） a prototype of the keyhole and the inverse Bremsstrahlung absorption coefficient in the keyhole plasma are obtained from the experiments;2） instead of using a parallel incident beam,a focused laser beam with real Gaussian intensity distribution is implemented;3） both Fresnel absorption and inverse Bremsstrahlung absorption during multiple reflections are considered.The calculation results show that the distribution of absorbed laser intensity by the keyhole wall is not uniform.The maximum laser energy is absorbed by the bottom of the keyhole,although no rays irradiate directly onto the bottom.According to analysis of beam focusing characteristics,the location of the focal plane plays a more important role in the laser energy absorption by the front wall than by the rear wall.

小剂量氯胺酮复合七氟醚麻醉用于小儿短小手术的临床观察

目的观察小剂量氯胺酮复合七氟醚麻醉在小儿短小手术中应用的安全性及可行性。方法选择2012年8月至2014年3月择期行短小手术的患儿80例,将其分成A、B两组,各40例;A组采用七氟醚复合氯胺酮麻醉,B组采用单纯氯胺酮麻醉。麻醉过程中监测两组患儿入室安静后5 min(T1)、麻醉诱导后(T2)、切皮时即刻(T3)、术毕(T4)呼吸循环相关参数,记录不良反应、苏醒时间及氯胺酮用量。结果 A组患儿氯胺酮用量明显少于B组,术后苏醒时间明显短于B组,各种不良反应发生率低于B组,差异均有统计学意义(P<0.05);B组T2、T3、T4时心率、平均动脉压上升均较A组显著,差异均有统计学意义(P<0.05)。结论小剂量氯胺酮复合七氟醚麻醉用于小儿短小手术比单纯氯胺酮麻醉更平稳,用药更少,苏醒更快,不良反应更少,值得临床推广应用。

Reproduction and biological characteristic of Chouioia cunea

Chouioia cunea Yang is a natural enemy for many kinds of crop and forest pests, particularly for Hyphantria cunea Drary, which is an international quarantine pest. The experiment of rearing Chouioia cunea with Tussah (Silkworm) pupa were carried out by 搕hree-cut method?and inoculating method. The results showed that three-cut method is effective way for breed-ing Chouioia cunea, with a parasitical rate of 95%. The biological characteristics and the life cycle of Chouioia cunea were ob-served and described and more hosts of Chouioia cunea were found. 揟hree-cut method?as a new technique of rearing Chouioia cunea has been put into practice.

LOW-THRUST ORBIT TRANSFER BY COMBINING GENETIC ALGORITHM WITH REFINED Q-LAW METHOD

A new orbit transfer method is presented by combining the genetic algorithm（GA）with the refined Q-law method.Considering the energy consumption,the relative thrust efficiency is introduced as a threshold deciding whether to thrust or coast.GA is used to achieve the global time-optimal orbit transfer.The trajectory optimization problem is transformed into the constraint parameter optimization problem,thus the nonlinear two-point boundary value problem is avoided.The refined Q-law method integrated with the fuzzy logic control is adopted for the end course,the vibration is avoided and the high precision is achieved.The numerical simulation of satellite orbit transfer is implemented.Results show that the new method can achieve the time-optimal orbit transfer and the low energy consumption,thus improving the transfer precision.

Nonlinear Auto-Companding Method for Behavioral Modeling of Switched-Current Circuits

A wavelet collocation method with nonlinear auto companding is proposed for behavioral modeling of switched current circuits.The companding function is automatically constructed according to the initial error distribution obtained through approximating the input output function of the SI circuit by conventional wavelet collocation method.In practical applications,the proposed method is a general purpose approach,by which both the small signal effect and the large signal effect are modeled in a unified formulation to ease the process of modeling and simulation.Compared with the published modeling approaches,the proposed nonlinear auto companding method works more efficiently not only in controlling the error distribution but also in reducing the modeling errors.To demonstrate the promising features of the proposed method,several SI circuits are employed as examples to be modeled and simulated.

Method for integer ambiguity resolution in GPS network RTK

A method for integer ambiguity resolution in the global positioning system （GPS） multi-reference station network real time kinematic （RTK） is proposed. First, the barycenter of the triangle of reference stations for ambiguity resolution is taken as a reference point. The satellite which has the largest elevation angle with the reference point is selected as a reference satellite. The parameters for constructing the weight matrix of carrier phase observation and the criteria for checking the correctness of integer ambiguity resolution of a network are obtained. Then, the wide ambiguity is calculated by a linear combination method of dualband observation. And the LI ambiguity is obtained by a nonionosphere combination method. The Kalman filter is introduced to refine the floating-point solution of ambiguity and estimate the real-time tropospheric delay. Finally, the cofactor matrix of ambiguity is de-correlated by Z-transformation to reduce the searching space of the integer ambiguity solution and improve the efficiency of the least-squares ambiguity decorrelation adjustment （LAMBDA） algorithm. The experimental results show that this method can reliably obtain the integer ambiguity solution among multi-reference stations with 40 epochs.

MRI Differentiation Diagnosis of Occupying Lesions in Cerebellopontine Angle Area

Objective： To explore the value of magnetic resonance imaging （MRI） in diagnosis of occupying lesions in cerebellopontine angle area. Methods： MRI records of 78 patients with pathologically confirmed occupied lesions in cerebellopontine angle area were analyzed. Results： Of the 78 cases, 48 （61.5%） were unilateral acoustic neuroma, 5 （6.4%） were bilateral acoustic neuroma, 12 （15.4%） were meningioma, 4 （5.1%） were trigeminal neuroma, 3 （3.8%） were lipoma, 2 （2.6%） were melanoma, and 1 （1.3%） was medulloblastoma. According to the anatomic site, tumor lesion character, and MRI signal, the majority of cerebellopontine angle area tumors were diagnosed accurately. Conclusion： MRI plays an important role in diagnosis of occupying lesions in cerebellopontine angle area.

Phase field method simulation of faceted dendrite growth with arbitrary symmetries

A numerical simulation based on a regularized phase field model is developed to describe faceted dendrite growth morphology. The effects of mesh grid, anisotropy, supersaturation and fold symmetry on dendrite growth morphology were investigated, respectively. These results indicate that the nucleus grows into a hexagonal symmetry faceted dendrite. When the mesh grid is above 640×640, the size has no much effect on the shape. With the increase in the anisotropy value, the tip velocities of faceted dendrite increase and reach a balance value, and then decrease gradually. With the increase in the supersaturation value, crystal evolves from circle to the developed faceted dendrite morphology. Based on the Wulff theory and faceted symmetry morphology diagram, the proposed model was proved to be effective, and it can be generalized to arbitrary crystal symmetries.

Novel registration algorithm for 3-D images captured from multiple views of object surface

A novel algorithm of 3-D surface image registration is proposed. It makes use of the array information of 3-D points and takes vector/vertex-like features as the basis of the matching. That array information of 3-D points can be easily obtained when capturing original 3-D images. The iterative least-mean-squared （LMS） algorithm is applied to optimizing adaptively the transformation matrix parameters. These can effectively improve the registration performance and hurry up the matching process. Experimental results show that it can reach a good subjective impression on aligned 3-D images. Although the algorithm focuses primarily on the human head model, it can also be used for other objects with small modifications.

Short-term forecasting optimization algorithms for wind speed along Qinghai-Tibet railway based on different intelligent modeling theories

To protect trains against strong cross-wind along Qinghai-Tibet railway, a strong wind speed monitoring and warning system was developed. And to obtain high-precision wind speed short-term forecasting values for the system to make more accurate scheduling decision, two optimization algorithms were proposed. Using them to make calculative examples for actual wind speed time series from the 18th meteorological station, the results show that: the optimization algorithm based on wavelet analysis method and improved time series analysis method can attain high-precision multi-step forecasting values, the mean relative errors of one-step, three-step, five-step and ten-step forecasting are only 0.30%, 0.75%, 1.15% and 1.65%, respectively. The optimization algorithm based on wavelet analysis method and Kalman time series analysis method can obtain high-precision one-step forecasting values, the mean relative error of one-step forecasting is reduced by 61.67% to 0.115%. The two optimization algorithms both maintain the modeling simple character, and can attain prediction explicit equations after modeling calculation.

Evaluation of the possible slip surface of a highly heterogeneous rock slope using dynamic reduction method

A new method, the dynamic reduction method(DRM) combined with the strain-softening method, was applied to evaluate the possible slip surface of a highly heterogeneous rock slope of the Dagangshan hydropower station in Southwest China.In DRM, only the strength of the failure elements is reduced and the softening reduction factor K is adopted to calculate the strength parameters. The simulation results calculated by DRM show that the further slip surface on the right slope of the Dagangshan hydropower station is limited in the middle part of the slope, while both SRM(strength reduction method) and LEM(limit equilibrium method) predict a failure surface which extends upper and longer. The observations and analysis from the three recorded sliding events indicate that the failure mode predicted by DRM is more likely the scenario.The results in this study illustrate that for highly heterogeneous slopes with geological discontinuities in different length scales, the proposed DRM can provide a reliable prediction of the location of the slip surface.

How much helpful is the capsule endoscopy for the diagnosis of small bowel lesions?

AIM： To assess the practically usefulness and diagnostic yield of this new method in a group of patients with suspected small bowel lesions. METHODS： Capsule endoscopic （CE） examination by using M2A capsule endoscope TM （Given Imaging, Yoqneam, Israel） was performed in thirty nine patients （26 males, 13 females） with suspected small intestinal lesions. The composing of the patients was as follows： obscure gastrointestinal bleeding in twenty three patients, known Crohn＇s disease in 6 patients, in whom CE was used to evaluate the severity and extension of the diseases, chronic diarrhea in 8 patients, abdominal pain in one patient and malignancy in one patient with unknown origin. RESULTS： In two patients CE failed. Different abnormalities were revealed in 26 patients overall. Detection rate of abnormalities was highest among patients with obscure gastrointestinal bleeding and the source of bleeding was demonstrated in 17 of 23 patients with obscure bleeding （73.9%）. Entero-Behcet was diagnosed in two patients by CE as a source of obscure gastrointestinal bleeding. In 6 patients with known Crohn＇s disease, CE revealed better evaluation of the disease extension. In 3 of 8 （37.5%） patients with chronic diarrhea; CE revealed some mucosal abnormalities as the cause of chronic diarrhea. In a patient with unexplained abdominal pain and in a cancer patient with unknown origin, CE examination was normal. CONCLUSION： In our relatively small series, we found that capsule endoscopy is a useful diagnostic tool particularly in diagnosis of obscure gastrointestinal bleeding, chronic diarrhea and in estimating the extension of Crohn＇s disease.

Wavelet matrix transform for time-series similarity measurement

A time-series similarity measurement method based on wavelet and matrix transform was proposed,and its anti-noise ability,sensitivity and accuracy were discussed. The time-series sequences were compressed into wavelet subspace,and sample feature vector and orthogonal basics of sample time-series sequences were obtained by K-L transform. Then the inner product transform was carried out to project analyzed time-series sequence into orthogonal basics to gain analyzed feature vectors. The similarity was calculated between sample feature vector and analyzed feature vector by the Euclid distance. Taking fault wave of power electronic devices for example,the experimental results show that the proposed method has low dimension of feature vector,the anti-noise ability of proposed method is 30 times as large as that of plain wavelet method,the sensitivity of proposed method is 1/3 as large as that of plain wavelet method,and the accuracy of proposed method is higher than that of the wavelet singular value decomposition method. The proposed method can be applied in similarity matching and indexing for lager time series databases.

Matrix metalloproteinases contribute to kidney fibrosis in chronic kidney diseases

Matrix metalloproteinases(MMPs) are members of the neutral proteinase family. They were previously thought to be anti-fibrotic because of their ability to degrade and remodel of extracellular matrix. However, recent studies have shown that MMPs are implicated in initiation and progression of kidney fibrosis through tubular cell epithelial–mesenchymal transition(EMT) as well as activation of resident fibroblasts, endothelial-mesenchymal transition(Endo MT) and pericyte-myofibroblast transdifferentiation. Interstitial macrophage infiltration has also been shown to correlate with the severity of kidney fibrosis in various chronic kidney diseases. MMPs secreted by macrophages, especially MMP-9, hasbeen shown by us to be profibrotic by induction of tubular cells EMT. EMT is mainly induced by transforming growth factor-β(TGF-β). However, MMP-9 was found by us and others to be up-regulated by TGF-β1 in kidney tubular epithelial cells and secreted by activated macrophages, resulting in EMT and ultimately kidney fibrosis. Therefore, MMP-9 may serve as a potential therapeutic target to prevent kidney fibrosis in chronic kidney disease. This review, by a particular focus on EMT, seeks to provide a comprehensive understanding of MMPs, especially MMP-9, in kidney fibrosis.

Multi-loop Internal Model Controller Design Based on a Dynamic PLS Framework

In this paper,a multi-loop internal model control(IMC) scheme in conjunction with feed-forward strategy based on the dynamic partial least squares(DyPLS) framework is proposed.Unlike the traditional methods to decouple multi-input multi-output(MIMO) systems,the DyPLS framework automatically decomposes the MIMO process into a multi-loop system in the PLS subspace in the modeling stage.The dynamic filters with identical structure are used to build the dynamic PLS model,which retains the orthogonality among the latent variables.To address the model mismatch problem,an off-line least squares method is applied to obtain a set of optimal filter parameters in each latent space.Without losing the merits of model-based control,a simple and easy-tuned IMC structure is readily carried over to the dynamic PLS control framework.In addition,by projecting the measurable disturbance into the latent subspace,a multi-loop feed-forward control is yielded to achieve better performance for disturbance rejection.Simulation results of a distillation column are used to further demonstrate this new strategy outperforms conventional control schemes in servo behavior and disturbance rejection.

DNN-Based Speech Enhancement Using Soft Audible Noise Masking for Wind Noise Reduction

This paper presents a deep neural network(DNN)-based speech enhancement algorithm based on the soft audible noise masking for the single-channel wind noise reduction. To reduce the low-frequency residual noise, the psychoacoustic model is adopted to calculate the masking threshold from the estimated clean speech spectrum. The gain for noise suppression is obtained based on soft audible noise masking by comparing the estimated wind noise spectrum with the masking threshold. To deal with the abruptly time-varying noisy signals, two separate DNN models are utilized to estimate the spectra of clean speech and wind noise components. Experimental results on the subjective and objective quality tests show that the proposed algorithm achieves the better performance compared with the conventional DNN-based wind noise reduction method.