基于支持向量机的计算资源反馈调度

在实时控制系统(RTCS)中,计算资源一般受限。由于诸多不确定性因素和工作负载的动态变化,系统总是运行在不可预期的开放环境中。为保证系统的稳定运行,提出了一种基于支持向量机(SVM)的计算资源反馈调度方法。它周期性地监测系统计算资源,通过SVM在线预测,得到各个控制回路的下一个采样周期,从而实现系统计算资源的动态分配。通过仿真实验验证了该反馈调度方法的性能,并与理想情况、传统开环调度方法进行了比较,显示了其优越性。     

基于二代bandelets域HMT模型的图像分割

在小波域隐马尔可夫树模型(WHMT)图像分割算法的基础上,根据二代handelets系数的分布统计特性,提出一种新的bandelets域隐马树模型(BHMT)有监督图像分割算法.新算法试图利用二代bandelets能表示图像几何规律性的特点来改进分割效果.为证明算法的有效性,用合成纹理图像和真实的航空和遥感图像做分割实验,并和小波域隐马模型树分割算法作了对比,对合成纹理图像分割,用错分概率作为评价参数.实验结果表明,此方法错分概率低、分割效果理想.     

引入趋近律的功率因数校正滑模控制仿真研究

提出一种引入趋近律的滑模变结构控制(SMVSC)方法来实现有源功率因数校正(APFC),减少电流的谐波成分.SMVSC是一种解决非线性时变系统(如APFC系统)问题的良好办法,但是实际应用中SMVSC的"抖振"现象问题必须要解决好.引入趋近律来削弱APFC系统在滑模控制中的"抖振"现象,并且依此推导出了控制APFC系统中的功率开关的PWM占空比.采用Matlab平台进行仿真验证控制策略控制APFC电路几乎得到单位功率因数,同时超调减小、响应时间缩短.     

Residual Feature Attentional Fusion Network for Lightweight Chest CT Image Super-Resolution

The diagnosis of COVID-19 requires chest computed tomography (CT). High-resolution CT images can provide more diagnostic information to help doctors better diagnose the disease, so it is of clinical importance to study super-resolution (SR) algorithms applied to CT images to improve the resolution of CT images. However, most of the existing SR algorithms are studied based on natural images, which are not suitable for medical images; and most of these algorithms improve the reconstruction quality by increasing the network depth, which is not suitable for machines with limited resources. To alleviate these issues, we propose a residual feature attentional fusion network for lightweight chest CT image super-resolution (RFAFN). Specifically, we design a contextual feature extraction block (CFEB) that can extract CT image features more efficiently and accurately than ordinary residual blocks. In addition, we propose a feature-weighted cascading strategy (FWCS) based on attentional feature fusion blocks (AFFB) to utilize the high-frequency detail information extracted by CFEB as much as possible via selectively fusing adjacent level feature information. Finally, we suggest a global hierarchical feature fusion strategy (GHFFS), which can utilize the hierarchical features more effectively than dense concatenation by progressively aggregating the feature information at various levels. Numerous experiments show that our method performs better than most of the state-of-the-art (SOTA) methods on the COVID-19 chest CT dataset. In detail, the peak signal-to-noise ratio (PSNR) is 0.11 dB and 0.47 dB higher on CTtest1 and CTtest2 at SR compared to the suboptimal method, but the number of parameters and multi-adds are reduced by 22K and 0.43G, respectively. Our method can better recover chest CT image quality with fewer computational resources and effectively assist in COVID-19.     

Adaptive control for a class of nonlinear systems with time-varying delays in state and input

This paper is concerned with the adaptive stabilization problem of uncertain input delayed systems.A solution to this problem is given for a class of uncertain nonlinear systems with time-varying delays in both state and input.An adaptive asymptotically stabilizing controller,which can guarantee the stability of the closed-loop system and the convergence of the original system state,is designed by means of the Lyapunov-Krasovskii functional stability theory combined with linear matrix inequalities (LMIs) an...     

Dynamically modulated intensity interrogation scheme using waveguide coupled surface plasmon resonance sensors

An electro-optically modulated intensity interrogation method based on tunable waveguide coupled surface plasmon resonance sensors has been proposed. It has been theoretically and experimentally demonstrated that the proposed scheme can enable sensitive measurement of measurand variations. By modulating the refractive index in the waveguide layer, this interrogation method yields modulated signal whose amplitude is related to measurand's refractive index. This amplitude modulated signal offers a higher signal to noise ratio and eliminates additive noise in the sensor system. A preliminary investigation using saline buffers with different NaCl concentrations shows a resolution of 2.3 × 10^?6 refractive index unit by our approach. Resolution can be controlled by the amplitude of the applied modulation voltage and can be further enhanced by optimizing the device structure or improving the electro-optical (E-O) coefficient of the E-O material. This approach is simple, stable, and promising for low-cost or multi-channel SPR biosensor applications.     

基于GML的市政道路管理模型

区别于传统的道路数据模型,基于地理标记语言（GML）的市政道路模型面向市政道路组成元素管理。分析市政道路组成元素的特点以及管理部门对市政道路管理的需求,依据GML的核心模式,建立面向管理领域的市政道路设施的空间属性一体化模型,并对该道路模型的应用前景进行分析。     

基于似然比测试的语音激活检测算法

针对低信噪比和噪声变化情况下语音激活检测算法性能下降的问题,提出一种新的参数更新和取值算法。该算法采用Laplacian-Gaussian混合模型对带噪语音频谱的概率分布进行建模,模型参数从带噪语音中估计得到,噪声功率参数通过跟踪语音的音节间隙进行平滑。实验结果表明,该算法在-5dB信噪比情况下,可以达到95％以上的检测率,具有优越的跟踪性能。     

Multi-scale object-based measurement of arid plant community structure

The measurement of plant community structure provides an extensive understanding of its function, succession and ecological process. The detection of plant community boundary is rather a challenge despite in situ work. Recent advances in object-based image analysis (OBIA) and machine learning algorithms offer new opportunities to address this challenge. This study presents a multi-scale segmentation approach to accurately identify the boundaries of each vegetation and plant community for mapping plant community structure. Initially, a very high resolution (VHR) Worldview-2 image of a desert area is hierarchically segmented from scale parameter 2 to 500. Afterward, the peak values of the standard deviation of brightness and normalized difference vegetation index (NDVI) across the segmentation scales are detected to determine the optimal segmentation scales of homogeneous single plant and plant community boundaries. A multi-scale classification of vegetation characterization with features of multiple bands, NDVI, grey-level co-occurrence matrix (GLCM) entropy and shape index is performed to identify dryland vegetation types. Finally, the four vegetation structural features on the type, diversity, object size and shape are calculated within the plant community boundaries and composed to plant community structure categories. Comparing the results with the object fitting index (FI) of the reference data, the validation indicates that the optimal segmentations of tree, shrub and plant communities are consistent with the identified peak values.