Scaling parameters selection principle for the scaled unscented Kalman filter

The paper deals with the state estimation of the widely used scaled unscented Kalman filter（UKF）. In particular, the stress is laid on the scaling parameters selection principle for the scaled UKF. Several problems caused by recommended constant scaling parameters are highlighted. On the basis of the analyses, an effective scaled UKF is proposed with self-adaptive scaling parameters,which is easy to understand and implement in engineering. Two typical strong nonlinear examples are given and their simulation results show the effectiveness of the proposed principle and algorithm.

Gearbox Fault Diagnosis using Adaptive Zero Phase Time-varying Filter Based on Multi-scale Chirplet Sparse Signal Decomposition

When used for separating multi-component non-stationary signals, the adaptive time-varying filter(ATF) based on multi-scale chirplet sparse signal decomposition(MCSSD) generates phase shift and signal distortion. To overcome this drawback, the zero phase filter is introduced to the mentioned filter, and a fault diagnosis method for speed-changing gearbox is proposed. Firstly, the gear meshing frequency of each gearbox is estimated by chirplet path pursuit. Then, according to the estimated gear meshing frequencies, an adaptive zero phase time-varying filter(AZPTF) is designed to filter the original signal. Finally, the basis for fault diagnosis is acquired by the envelope order analysis to the filtered signal. The signal consisting of two time-varying amplitude modulation and frequency modulation(AM-FM) signals is respectively analyzed by ATF and AZPTF based on MCSSD. The simulation results show the variances between the original signals and the filtered signals yielded by AZPTF based on MCSSD are 13.67 and 41.14, which are far less than variances (323.45 and 482.86) between the original signals and the filtered signals obtained by ATF based on MCSSD. The experiment results on the vibration signals of gearboxes indicate that the vibration signals of the two speed-changing gearboxes installed on one foundation bed can be separated by AZPTF effectively. Based on the demodulation information of the vibration signal of each gearbox, the fault diagnosis can be implemented. Both simulation and experiment examples prove that the proposed filter can extract a mono-component time-varying AM-FM signal from the multi-component time-varying AM-FM signal without distortion.

Empirical mode decomposition using variable filtering with time scale calibrating

A novel and efficient method for decomposing a signal into a set of intrinsic mode functions （IMFs） and a trend is proposed. Unlike the original empirical mode decomposition （EMD）, which uses spline fits to extract variations from the signal by separating the local mean from the fluctuations in the decomposing process, this new method being proposed takes advantage of the theory of variable finite impulse response （FIR） filtering where filter coefficients and breakpoint frequencies can be adjusted to track any peak-to-peak time scale changes. The IMFs are results of a multiple variable frequency response FIR filtering when signals pass through the filters. Numerical examples validate that in contrast with the original EMD, the proposed method can fine-tune the frequency resolution and suppress the aliasing effectively.

COLOR-SCALE MORPHOLOGICAL FILTERS WITH MULTIPLE STRUCTURING ELEMENTS

With the wide use of color in many areas, the interest on the color perception and processing has been growing rapidly. An important topic in color image processing is the development of efficient tools capable of filtering images without blurring them and without changing their original chromatic contents. In this paper, a new technique reducing noise of color image is developed. A class of color-scale morphological operations is introduced, which extend mathematical morphology to color image processing, representing a color image as a vector function. The correlation between color components is utilized to perform noise removal. Color-scale morphological niters with multiple structuring elements (CSMF-MSEs) are proposed. Their properties are discussed and proved. Experimental results show that CSMF-MSEs are suitable and powerful to eliminate noise and preserve edges in color image because of efficient utilization of inherent correlation between color components, and they perform better than vector

模块环境下的filter-SQP用于过程优化

Sequential Quadratic Programming(SQP) is the most efficient algorithm for nonlinear optimization.But a penalty function is usually used for linear search,causing some problems.Filter-SQP developed by Roger Fletcher and Sven Leyffer avoids using penalty function.In the view of filter-SQP,NLP problem has two objectives,one is minimizing objective function,the other is satisfying the constraints.The concept of filter is proposed on the basis of these two objectives.In this paper flowsheet optimization using filter-SQP in modular simulator environment was studied.Infeasible path strategy was used and the constraint function was composed of tear stream equation,specific design and unsatisfied inequality constraint.When filter could not find a step as the starting point of the next iteration,in order to avoid algorithm failure three strategies were used.They were restarting strategy,converging recycle strategy and feasible path strategy.A successive scaling strategy was proposed for filter-SQP to improve the efficiency of optimization.A case study of process optimization with filter-SQP was very encouraging.

A TIME FILTERING SCHEME FOR THE SHORT RANGE CLIMATE PREDICTION MODEL PRODUCTS AND ITS REAL CASE ANALYSIS

A filtering / extracting scheme for various timescale processes in short range climate model out-put is established by using the scale scattering method. And the climatological meanings as well as the impor-tance of the filtered series are discussed. In the latter part of work, the effectiveness of the filtering method and the performance of the prediction model are analyzed through a real case.

A Quaternion Scaled Unscented Kalman Estimator for Inertial Navigation States Determination Using INS/GPS/Magnetometer Fusion

This Inertial Navigation System (INS), Global Positioning System (GPS) and fluxgate magnetometer technologies have been widely used in a variety of positioning and navigation applications. In this paper, a low cost solid state INS/GPS/Magnetometer integrated navigation system has been developed that incorporates measurements from an Inertial Navigation System (INS), Global Positioning System (GPS) and fluxgate magnetometer (Mag.) to provide a reliable complete navigation solution at a high output rate. The body attitude estimates, especially the heading angle, are fundamental challenges in a navigation system. Therefore targeting accurate attitude estimation is considered a significant contribution to the overall navigation error. A better estimation of the body attitude estimates leads to more accurate position and velocity estimation. For that end, the aim of this research is to exploit the magnetometer and accelerometer data in the attitude estimation technique. In this paper, a Scaled Unscented Kalman Filter (SUKF) based on the quaternion concept is designed for the INS/GPS/Mag integrated navigation system under large attitude error conditions. Simulation and experimental results indicate a satisfactory performance of the newly developed model.

基于改进直方图均衡的SF6泄漏区域增强算法

SF6气体红外成像易受环境噪声影响,呈低对比度与低信噪比特性。导致现有算法无法自适应增强SF6泄漏区域和抑制高斯噪声。针对上述问题,提出一种基于改进HE的SF6泄漏区域增强算法。该算法首先采用单尺度Retinex处理原始SF6图像获得反射图像,然后利用引导滤波将反射图像分解为细节层和基本层,最后采用改进的直方图均衡化来自适应处理基本层,并将增强后的图像进行融合来获得最终的图像。实验结果表明:本文算法不仅能够自适应增强泄漏区域的对比度,而且具有良好边缘保持特性和抑制高斯噪声的性能。其增强效果优于现有的SF6红外图像增强算法。有效改善了SF6红外图像低对比度和低信噪比特性。

采用局部-全局区域重检测机制的无人机长期跟踪算法

为解决基础跟踪器面对遮挡和移出视野等长期跟踪场景时易出现跟踪失败等问题,提出了一种基于局部-全局区域重检测的无人机长期跟踪算法。设计了基础滤波器,将高置信度样本与其结合,并融入自适应时空正则化,解决了滤波器退化问题,提高了模型鲁棒性以及复杂场景下的性能;优化了滤波器更新策略,通过评价跟踪结果进行自适应更新;设计快速尺度滤波器,解决了跟踪过程中的尺度变化问题;设计了局部-全局区域重检测机制,跟踪失败时启动重检测器恢复跟踪目标,先完成局部区域重检测,若恢复跟踪失败,再利用全局区域重检测器继续恢复目标跟踪状态。实验结果表明:所提算法在UAV20L数据集上的精确度和准确率分别可达0.724和0.621,与基于时空正则化相关滤波器的跟踪算法(STRCF)相比分别提升了25.9%和20.6%,与同类主流算法相比,跟踪效果得到提升,证明了算法的有效性。

基于核相关滤波的深空目标跟踪算法

随着天基红外探测器接近目标,目标在焦平面的成像会由小变大,在近距离时以面目标形式成像,随着距离的接近,目标的尺度变化会越来越剧烈。针对经典核相关滤波算法无法应对尺度变化及遮挡的问题,论文提出了一种融合尺度滤波的核相关滤波跟踪算法。在位置滤波器预测目标位置之后,由尺度滤波器计算尺度响应以解决目标尺度变化问题。针对目标跟踪过程中的遮挡问题,提出了一种遮挡判别依据,遮挡情况下停止更新模板以实现稳定、鲁棒的跟踪。通过对比实验表明,该算法可以有效地应对尺度变化及遮挡情况,且帧率可以达到120 FPS,满足实时要求。

一种大尺度区域GNSS坐标时间序列自适应时空滤波方法

提出一种GNSS坐标时间序列自适应时空滤波方法,在规定阈值下将滤波区域自适应分为若干个子区域,进行共模误差的提取和去除。对陆态网184个GNSS站点垂向坐标序列进行时空滤波,3组随机实验中,自适应PCA时空滤波后的站点序列平均RMS值减少约39.7%、38.4%和39.7%,且优于整体PCA滤波。进一步分析滤波前后站点噪声特性变化,结果显示,相比于整体PCA滤波,自适应滤波方法中站点残差序列幂律噪声减少约17.8%。

GNSS/MEMS IMU车载组合导航中IMU比例因子误差的影响分析

从系统状态模型出发,分析比例因子误差对组合导航精度和计算量的影响,同时基于车载运动的特点分析比例因子误差的观测性,提出一种仅保留航向陀螺仪和水平加速度计比例因子误差的降维状态模型。实验结果表明,当比例因子误差大于6×10^(-3)时,增广比例因子误差有助于提高导航精度,但计算量增加约170%;降维模型能够达到高维模型的导航精度,与不增广比例因子误差相比,计算量仅增加约70%。

基于非线性各向异性滤波的图像特征匹配算法

图像特征匹配是增强现实系统中的关键技术,匹配精度是提升特征匹配性能的关键。提出了一种多尺度特征匹配加强算法(I-AKAZE),通过对非线性各向异性滤波过程中传导函数的改进,减缓图像梯度值大的区域非线性扩散速度,极大程度地保留了匹配图像的边缘特征;同时,结合改进的非线性量化加速稳健特征描述符(NLG-SURF),提高了描述符的识别率。实验结果表明I-AKAZE算法在Mikolajczyk数据集上的可重复性得分相比目前先进的AKAZE算法有着大幅度提升,对应的特征描述符的平均识别率提升8.4%,并且运行速度比经典的SIFT算法快约19%,算法整体在检测和描述阶段上的性能都有提升。

Characterization and Laboratory-Scale Treatment of Municipal Drainage Wastewater of Khulna, Bangladesh

The study was conducted to characterize and perform laboratory-scale treatment of municipal drainage wastewater of Khulna, Bangladesh. Wastewater samples were collected from three different points of existing urban drain outlets into the Mayur River around Khulna. Laboratory testing shows the BOD5 and COD concentration of wastewater samples varied from 57 - 226 mg/l and 320 - 435 mg/l, respectively, and the total dissolved solids ranged from 1800 - 2525 mg/l. Therefore, a laboratory-scale treatment technology was developed to treat this wastewater. Treatment technologies adopted were primary sedimentation, followed by aeration, chemical precipitation and filtration. In treated wastewater, BOD5, COD and TDS were found to be in the range of 40 - 115 mg/l, 160 - 256 mg/l and 1356 - 1500 mg/l, respectively. These test results suggest that the performance of laboratory-scale treatment plant was not adequate to fulfil the acceptable limit (ECR’97) for safe disposal into surface water bodies. Due to poor quality of effluents, modification of laboratory-scale treatment plant was made by an activated sludge process followed by granular media filtration. The final BOD5, COD TDS and TSS concentration of effluents was found to be 1.38 - 9.8 mg/l, 32 - 192 mg/l, 590 - 1667 mg/l, and 35 - 95 mg/l respectively, which satisfy ECR’97 standard limits for safe disposal into inland water bodies.

Analysis on the Meso-scale System Activity and the Structure Characteristics of A Short-time Rainstorm

By using the rainfall data in the regional automatic station,FY-2E satellite data,NCEP reanalysis data,the evolution features and the structure characteristics of a meso-scale convective system(MCS) which happened on May 6 in 2010 in Loudi City of Hunan Province were analyzed.The results showed that MCS was the important influence system for the generation and development of strong precipitation.The equivalent blackbody brightness temperature(TBB) field of satellite inversion could directly reflect the convective activity of cumulus,the precipitation distribution and the intensity characteristics in the rainstorm process.TBB low value belt had the good corresponding relationship with the rainstorm falling zone.The disturbance flow field and the height field which passed Barnes band-pass wave filtering represented that there existed the obvious high-layer anticyclonic circulation and the low-layer cyclonic circulation near the rainstorm zone.The divergence in the high layer and the convergence in the low layer enhanced the occurrence and development of MCS.In addition,the disturbance temperature field revealed the main source of energy which the occurrence and development of strong convective weather needed.

大尺度磁性目标跟踪方法研究

针对磁定位中由于目标尺度不可忽略导致的观测模型不匹配问题,提出一种仅需2个三轴磁传感器测量数据,且能够自适应目标尺度的滤波估值型实时磁定位算法。该方法基于最大似然选择思想,建立了不同个数磁偶极子构成的均匀线列阵模型作为候选观测模型;推导了一种适用于高维非线性动态参数估计的随机高斯流滤波算法,克服了大尺度目标模型高维度和强非线性导致的性能下降问题,提高了实时定位精度。同时,采用该算法对各个偶极子线阵模型分别进行滤波解算,选取最大似然值对应的滤波结果作为当前时刻估计值,实现了对目标尺度的自适应。通过双层壳体潜艇有限元磁场仿真模型数据对算法的有效性进行了验证,结果表明,相比现有方法,该算法估值更为准确,能够满足精确实时定位目标的需求,且具有更好的收敛性能。

基于滤波器剪枝的多尺度压缩感知图像重构

为了解决低采样率下单尺度采样的压缩感知重构图像纹理细节模糊问题,同时达到使重构网络轻量化的目的,提出了一种基于滤波器剪枝的多尺度压缩感知图像重构网络.采样阶段,通过卷积来模拟图像的线性分解,融合输入图像和不同尺度的分解特征后完成多尺度采样,得到压缩感知测量值.重构阶段,设计了一种基于坐标注意力的多尺度空洞残差模块,将位置信息嵌入通道注意力中,增强网络特征学习的能力.同时通过计算特征图的熵来判断滤波器的重要性,剪除重要性较低的滤波器,达到压缩模型的目的.在DIV2K、Set5、BSDS68和Urban100等数据集上进行训练及测试.实验结果表明,所提算法在峰值信噪比(Peak Signalto-Noise Ratio,PSNR)、结构相似性(Structural Similarity,SSIM)和图像视觉效果上均有提升.其中,采样率为4%,测试集为Set14时,与CSNet+和FSOINet相比,所提算法将重构图像的PSNR分别提高了4.17 dB和2.39 dB,纹理细节更加清晰.在重构效果略微降低的前提下,得到更轻量化的模型,提升了重构速度.

融合DSST-KCF算法的城市低空反无人机自动图像追踪系统

面向复杂城市低空环境中的“低、慢、小”无人机的精准快速跟踪难题,提出了融合DSST-KCF算法的反无人机自动追踪系统。该系统由双光图像采集单元和显示控制单元两部分组成;双光图像采集单元采集低空图像,然后由融合DSST-KCF算法对图像进行实时处理;融合DSST-KCF跟踪算法包括位置滤波器和尺度滤波器,其中位置滤波器采用循环矩阵形成的数据集进行训练,然后预测出下一帧目标位置;尺度滤波器从目标位置中心提取出33种不同尺度下的特征进行训练,然后确定跟踪框的尺寸;最后采用DSST-KCF-云台控制策略,使无人机保持在视场中心附近实现有效跟踪。在校园环境中,利用本系统对空中无人机进行实时跟踪测试,测试结果表明,改进后算法相比原KCF算法检测成功率平均提升了31.51%,跟踪精度平均提升了43%,适用于小型运动目标的跟踪。

无线传感器网络的鲁棒分布式估计算法

为了提高无线传感器网络在非高斯噪声环境下的分布式估计性能,提出了一种基于半二次准则的变尺度扩散式自适应滤算法,该算法采用了一种凸代价函数的半二次准则(half-quadratic criterion,HQC),使算法可以更加有效地寻找最优解、具有更快的收敛速度和更高的滤波精度,并且对脉冲噪声具有更强的鲁棒性.另外,通过引入变尺度因子来平衡和进一步提高算法的收敛速度与稳态误差性能.文中对算法的性能进行了理论分析和仿真对比,结果表明在不同的非高斯噪声环境下,所提算法对未知系统具有更高的辨识效率,其收敛性能和稳态误差性能都优于其他的扩散式自适应滤波算法.