深度多模态不确定度的短视频事件检测方法

随着短视频的快速发展,短视频事件检测任务受到越来越多的关注。现有短视频事件检测研究普遍采用深度神经网络来获得确定的检测结果,但是网络忽略了不确定度的影响从而导致错误的预测结果也会产生过度置信的决策。为了解决上述问题,本文提出了一个深度多模态不确定度网络的短视频事件检测方法。首先,该方法在传统域分离网络中嵌入变分层,用来获得预测分布;然后,将视觉模态信息和音频模态信息输入到网络中,利用该方法所构建的独立性和相关性损失可以获得包含不确定度的音频模态共、私有域预测分布以及视觉模态共、私有域预测分布;最后,提出了一个不确定度判别法则用来筛选4个域的预测分布,从而得到最终的预测结果。在公开数据集(UCF-101与HMDB51)和新构建的短视频事件检测数据集上进行了实验。实验结果表明,面对不同的深度分类方法以及不同的数据集,本文方法不仅有着更高的分类准确率,还可以对输出结果进行不确定度估计,针对音频的干扰也具有较强的鲁棒性。

一种不确定模态缺失的多模态对抗虚假新闻检测框架

当前虚假新闻的检测方法已经从传统的单一模态数据分析转向了多模态数据融合技术的应用。然而现有的多模态虚假新闻检测方法未充分考虑到现实场景中可能存在的模态缺失问题。模态的缺失可能会导致无法完整地抽取和融合新闻特征,从而降低模型的性能。该文提出一种新的虚假新闻检测框架,该框架在多模态特征学习中,加入了一个模态鉴别器,其在与特征生成器进行对抗训练的过程中学习不同模态组合之间的迁移特征,实现了在不确定模态缺失场景下的虚假新闻检测。通过在真实数据集上进行实验证明,该文所提出的框架在不确定模态缺失场景下优于最新的多模态虚假新闻检测方法。

基于变分模态分解的风机滚动轴承早期故障诊断

针对海上风电机组滚动轴承故障多且早期故障特征难以提取的问题,提出了一种基于变分模态分解的滚动轴承故障诊断方法。从频率方面研究了模态分量个数对信号特征信息的影响,提出故障特征信息提取时确定分解个数的一般原则,据此确定滚动轴承早期故障振动信号的分解个数并得到若干模态分量,从中筛选出最佳模态分量进行包络解调分析,最终通过对比包络谱特征频率对滚动轴承进行早期故障诊断。

基于两阶段改进的FCM法的模态参数自动识别

为解决基于稳定图的模态参数自动识别中存在人工干预、参数设定不统一的问题,提出两阶段改进的稳定图自动识别方法.首先,将模态参数不确定度指标引入第一阶段虚假模态参数剔除过程,实现最大程度虚假模态参数剔除;然后,基于改进的FCM算法,采用迭代策略计算不同聚类数目的隶属度矩阵,构造累积邻接矩阵,并结合图切割算法解析累积邻接矩阵,自动确定最佳聚类数目,实现稳定图自动识别;最后,将提出的改进方法运用在Z24桥Benchmark模型和悬索桥实测数据上,验证了所提方法的可行性.研究结果表明:模态参数不确定度相较于传统指标对虚假模态具有更强的辨别能力;基于改进的FCM算法不需要任何人工调整的参数就能自动识别稳定轴,且具有较强的鲁棒性.改进算法在默认参数下准确识别得到Z24桥和悬索桥的模态参数,表明提出的改进算法可以用于桥梁健康监测的模态参数自动识别过程.

FSDD运行模态参数识别方法中不确定性的计算

运行模态参数识别是一种只利用响应数据进行模态分析的方法.环境激励条件下获得的识别结果会存在随机不确定性,这种不确定程度可以通过模态参数的统计特性进行描述评价.以运行模态分析中的频域空间域分解方法(FSDD)为研究对象,通过对整个模态识别算法进行摄动分析,确定出模态参数对测量误差的灵敏度,实现了模态参数方差的计算.计算过程分为四步:第一步,响应信号功率谱矩阵中噪声方差的估计;第二步,增强功率谱方差的计算;第三步,极点协方差的计算;第四步,模态参数方差的计算.最后通过一个4自由度仿真算例与张家港桥梁实测算例对所给出的方法进行了验证与应用.

Sign Flip in Giorgi-Pasquale-Paganelli Model

Taking advantage of our improved gates as well as the Giorgi-Pasquale-Paganelli （GPP） nondestructive gate [Phys. Rev. A 70 （2004） 022319], the sign of any term of the initial state in the GPP model can be flipped. The success possibility of the sign flip of odd number of terms is probabilistic （11/4）, while the success possibility of the sign flip of even number of terms is deterministic （1）.

Uncertain Modal Analysis of Unmanned Aircraft Composite Landing Gear

Through taking uncertain mechanical parameters of composites into consideration,this paper carries out uncertain modal analysis for an unmanned aircraft landing gear.By describing correlated multi-dimensional mechanical parameters as a convex polyhedral model,the modal analysis problem of a composite landing gear is transferred into a linear fractional programming(LFR)eigenvalue solution problem.As a consequent,the extreme-point algorithm is proposed to estimate lower and upper bounds of eigenvalues,namely the exact results of eigenvalues can be easily obtained at the extreme-point locations of the convex polyhedral model.The simulation results show that the proposed model and algorithm can play an important role in the eigenvalue solution problem and possess valuable engineering significance.It will be a powerful and effective tool for further vibration analysis for the landing gear.

Adaptive robust motion trajectory tracking control of pneumatic cylinders

High-accuracy motion trajectory tracking control of a pneumatic cylinder driven by a proportional directional control valve was considered. A mathematical model of the system was developed firstly. Due to the time-varying friction force in the cylinder, unmodeled dynamics, and unknown disturbances, there exist large extent of parametric uncertainties and rather severe uncertain nonlinearities in the pneumatic system. To deal with these uncertainties effectively, an adaptive robust controller was constructed in this work. The proposed controller employs on-line recursive least squares estimation(RLSE) to reduce the extent of parametric uncertainties, and utilizes the sliding mode control method to attenuate the effects of parameter estimation errors, unmodeled dynamics and disturbances. Therefore, a prescribed motion tracking transient performance and final tracking accuracy can be guaranteed. Since the system model uncertainties are unmatched, the recursive backstepping design technology was applied. In order to solve the conflicts between the sliding mode control design and the adaptive control design, the projection mapping was used to condition the RLSE algorithm so that the parameter estimates are kept within a known bounded convex set. Extensive experimental results were presented to illustrate the excellent achievable performance of the proposed controller and performance robustness to the load variation and sudden disturbance.

Fuzzy finite element model updating of bridges by considering the uncertainty of the measured modal parameters

It is significant to consider the effect of uncertainty of the measured modal parameters on the updated finite element(FE) model,especially for updating the FE model of practical bridges,since the uncertainty of the measured modal parameters cannot be ignored owing to the application of output-only identification method and the existence of the measured noise.A reasonable method is to define the objective of the FE model updating as the statistical property of the measured modal parameters obtained by conducting couples of identical modal tests,however,it is usually impossible to implement repeated modal test due to the limit of practical situation and economic reason.In this study,a method based on fuzzy finite element(FFM) was proposed in order to consider the effect of the uncertainty of the measured modal parameters on the updated FE model by using the results of a single modal test.The updating parameters of bridges were deemed as fuzzy variables,and then the fuzzification of objective of the FE model updating was proposed to consider the uncertainty of the measured modal parameters.Finally,the effectiveness of the proposed method was verified by updating the FE model of a practical bridge with the measured modal parameters.

Model updating based on incomplete modal data

The objective of model updating is to improve the accuracy of a dynamic model based on the correlation between the measured data and the analytical (finite element) model. In this paper, we intend to update the mass and stiffness matrices of an analytical model when only modal frequencies or spatially incomplete modal data are available. While the proposed method is systematic in nature, it also preserves the initial configuration of the analytical model, and physical equality and/or inequality constraints can be easily incorporated into the solution procedure. Numerical examples associated with a simple 5-DoF (degree of freedom) mass-spring system are chosen to illustrate the detailed procedure and the effectiveness of the proposed method. Numerical scenarios ranging from the updating for stiffness terms only to that for all mass and stiffness terms based on various kinds of incomplete modal data are studied. The obtained model updating results are excellent when the measured modal data are noise-free. Uncertainty studies are also conducted based on simulations of corrupted modal data, but a thorough theoretical analysis of the noise effect on the proposed method is still needed.

Hybrid analytical resolution approach based on ambiguity function for attitude determination

When satellite navigation receivers are equipped with multiple antennas, they can deliver attitude information. In previous researches, carrier phase differencing measurement equations were built in the earth-centered, earth-fixed (ECEF) co- ordinate, and attitude angles could be obtained through the rotation matrix between the body frame (BF) and the local level frame (LLF). Different from the conventional methods, a hybrid algorithm is presented to resolve attitude parameters utilizing the single differencing (SD) carrier phase equations established in LLF. Assuming that the cycle integer ambiguity is known, the measure- ment equations have attitude analytical resolutions by using simultaneous single difference equations for two in-view satellites. In addition, the algorithm is capable of reducing the search integer space into countable 2D discrete points and the ambiguity function method (AFM) resolves the ambiguity function within the analytical solutions space. In the case of frequency division multiple access (FDMA) for the Russian Global Orbiting Navigation Satellite System (GLONASS), a receiver clock bias estimation is employed to evaluate its carrier phase. An evaluating variable and a weighted factor are introduced to assess the integer ambiguity initialization. By static and dynamic ground experiments, the results show that the proposed approach is effective, with enough accuracy and low computation. It can satisfy attitude determination in cases of GPS alone and combined with GLONASS.

Characteristic model-based consensus of networked heterogeneous robotic manipulators with dynamic uncertainties

In this paper, we address the characteristic model-based discrete-time consensus problem of networked robotic manipulators with dynamic uncertainties. The research objective is to achieve joint-position consensus of multiple robotic agents interconnected on directed graphs containing a spanning tree. A novel characteristic model-based distributed adaptive control scenario is proposed with a state-relied projection estimation law and a characteristic model-based distributed controller. The performance analysis is also unfolded where the uniform ultimate boundedness(UUB) of consensus errors is derived by resorting to the discrete-time-domain stability analysis tool and the graph theory. Finally, numerical simulations illustrate the effectiveness of the proposed theoretical strategy.

Uncertainty and Sensitivity Analyses for Modeling Long-Term Soil Organic Carbon Dynamics of Paddy Soils Under Different Climate-Soil-Management Combinations

Reporting modeling results with uncertainty information can benefit decision making by decreasing the extent that variability exerts a disproportionate influence on the options selected. For making decisions with more confidence, the uncertainty interval should be as narrow as possible. Here, the soil organic carbon （SOC） dynamics of the major paddy soil subgroup from 4 different paddy field regions of China （located in 4 counties under different climate-soil-management combinations） were modeled using the DeNitrification- DeComposition （DNDC） model for the period from 1980 to 2008. Uncertainty intervals associated with the SOC dynamics for these 4 subgroups were estimated by a long-term global sensitivity and uncertainty analysis （i. e., the Sobolt method）, and their sensitivities to 7 influential factors were quantified using the total effect sensitivity index. The results, modeled with high confidence, indicated that in the past 29 years, the studied paddy soils in Xinxing, Yixing, and Zhongjiang counties were carbon （C） sinks, while the paddy soil in Helong County was a C source. The 3 C sinks sequestered 12.2 （5.4, 19.6）, 17.1 （8.9, 25.0）, and 16.9 （-1.2, 33.6） t C ha-1 （values in the parentheses are the 5th and 95th percentiles, respectively）. Conversely, the C source had a loss of -5.4 （-14.2, 0.06） t C ha-1 in the past 29 years. The 7 factors, which changed with the climate-soil-management context, exhibited variable influences on modeled SOC. Measures with potential to conserve or sequestrate more C into paddy soils, such as incorporating more crop residues into soils and reducing chemical fertilizer application rates, were recommended for specific soils based on the sensitivity analysis results.