Retrieval of nighttime aerosol optical depth by simultaneous consideration of artificial and natural light sources.

Aerosol Optical Depth (AOD) is a critical optical parameter that quantifies the degree of light attenuation by aerosols and serves as a fundamental indicator of atmospheric quality. Therefore, accurate quantification and retrieval of AOD is crucial for relevant studies. However, current satellite-based AOD retrieval algorithms suffer from inapplicability under low-light conditions, limiting the development of nighttime AOD retrieval. Under this context, we proposed a novel algorithm, namely Simultaneous Consideration of Artificial and Natural light Sources (SCANS), to obtain nighttime AOD. The core of the SCANS algorithm is considering the synergy of both the natural and artificial light sources to obtain nighttime AOD by integrating atmospheric radiative transfer simulation into an extinction method and performing multiple iterations. SCANS was applied to the Visible Infrared Imaging Radiometer Suite (VIIRS) Day/Night Band (DNB) and the retrieved nighttime AOD was validated with in-situ measurements from five AERONET sites. Results indicate that the Mean Bias Errors (MBEs) of the retrieved nighttime AOD range from 0.0 to 0.08 and the corresponding Root Mean Square Errors (RMSEs) range from 0.11 to 0.17, which exhibit better accuracy than that of the nighttime MERRA-2 AOD. We further compared the retrieved nighttime AOD with the corresponding Air Quality Index (AQI) measurements at six environment monitoring stations and obtained high correlation coefficients (i.e., ranging from 0.733 to 0.940), indicating SCANS's reliability and high accuracy. The proposed SCANS algorithm can effectively obtain nighttime AOD with high quality, thereby advancing research on the diurnal variation of crucial Earth's key elements.

Hierarchical attention network for multivariate time series long-term forecasting.

Multivariate time series long-term forecasting has always been the subject of research in various fields such as economics, finance, and traffic. In recent years, attention-based recurrent neural networks (RNNs) have received attention due to their ability of reducing error accumulation. However, the existing attention-based RNNs fail to eliminate the negative influence of irrelevant factors on prediction, and ignore the conflict between exogenous factors and target factor. To tackle these problems, we propose a novel Hierarchical Attention Network (HANet) for multivariate time series long-term forecasting. At first, HANet designs a factor-aware attention network (FAN) and uses it as the first component of the encoder. FAN weakens the negative impact of irrelevant exogenous factors on predictions by assigning small weights to them. Then HANet proposes a multi-modal fusion network (MFN) as the second component of the encoder. MFN employs a specially designed multi-modal fusion gate to adaptively select how much information about the expression of current time come from target and exogenous factors. Experiments on two real-world datasets reveal that HANet not only outperforms state-of-the-art methods, but also provides interpretability for prediction.

An effective behavior recognition method in the video session using convolutional neural network.

In order to further improve the accuracy of the video-based behavior recognition method, an effective behavior recognition method in the video session using convolutional neural network is proposed. Specifically, by adding the target detection phase before the behavior recognition algorithm, the body region in the video can be accurately extracted to reduce the interference of redundant and unnecessary background noises, and at the same time, the inappropriate images can be replaced, which has reached the role of balance background trade-off, and finally, the neural network can learn the human behavior information with emphasis. By adding fragmentation and stochastic sampling, the long-time time-domain modeling of the whole video session can be established, so that the model can obtain video-level expression ability. Finally, the improved loss function is used for behavior recognition to solve the problem of classification difficulty and possible sample imbalance. In addition, we conducted the hyperparametric experiment, the ablation experiment and the contrast experiment on different open source and benchmark datasets. Compared with other commonly used behavior recognition algorithms, the experimental results verify the effectiveness of the proposed method. In addition, the related deep learning-based methods used in behavior recognition are reviewed at the beginning of this paper, and the challenges in behavior recognition and future research directions are prospected at the end of this paper, which will undoubtedly play a double role in the work of later researchers.

An observer-based adaptive fault-tolerant control for hypersonic vehicle with unexpected centroid shift and input saturation.

This paper makes an investigation on the fault-tolerant control (FTC) problem for a hypersonic reentry vehicle (HSV) in the coexistence of unknown movement of center-of-mass, system input constraint and failure of actuator. Firstly, the dynamics of HSV's attitude system with the unknown factors mentioned above are developed to illustrate the particularity of the researched topic. The influences of unexpected centroid shift on an FTC design can be summarized into the following three parts: unknown system uncertainties, eccentric torque as well as changing system moment of inertia matrix, which are coupled and unknown. Secondly, due to the difficulty in decoupling and estimating these influences (embodied in the output states of the system) one by one, it is the attitude system states observer that is proposed to estimate those detrimental unknown effects. The designed observer is consisted of an adaptive fault observer and an adaptive sliding-mode observer, supporting an innovative adaptive FTC scheme free from the variation of inverse matrix that might be singular due to an unexpected centroid shift. This fault-tolerant controller established in the estimated system states is derived by utilizing the above mentioned observer and adaptive backstepping control in conjunction with adaptive auxiliary compensation systems to handle the system input saturation. Moreover, the convergence of attitude tracking error and the boundedness of all closed-loop signals are achieved in the light of Lyapunov stability theory and boundedness analysis. Ultimately, simulation results are delivered to demonstrate the effectiveness of the proposed FTC scheme.