Abstract Photovoltaic (PV) is characterized by random and intermittent. As increasing popularity of PV, it makes PV power prediction increasingly significant for efficiency and stability of the power grid. At present, prediction models of PV power based on deep learning show superior performance, but they ignore the interdependent mechanism of prediction error along the input characteristics of the neural network. This paper proposed a self-attention mechanism (SAM)-based hybrid one-dimensional convolutional neural network (1DCNN) and long short-term memory (LSTM) combined method (named 1DCNN-LSTM-SAM). In the proposed model, SAM redistributes the neural weights in 1DCNN-LSTM, and then 1DCNN-LSTM further extracts the space-time information of effective PV power. The polysilicon PV arrays data in Australia are employed to test and verify the proposed model and other five competition models. The results show that the application of SAM to 1DCNN-LSTM improves the ability to capture the global dependence between inputs and outputs in the learning process and the long-distance dependence of its sequence. In addition, mean absolute percentage error of the 1DCNN-LSTM-SAM under sunny day, partially cloudy day, and cloudy day weather types has increased by 24.2%, 14.4%, and 18.3%, respectively, compared with the best model among the five models. Furthermore, the weight distribution mechanism of self-attention to the back end of LSTM was analyzed quantitatively and the superiority of SAM was verified.
The explosive growth of digital images in video surveillance and social media has led to the significant need for efficient search of persons of interest in law enforcement and forensic applications. Despite tremendous progress in primary biometric traits (e.g., face and fingerprint) based person identification, a single biometric trait alone can not meet the desired recognition accuracy in forensic scenarios. Tattoos, as one of the important soft biometric traits, have been found to be valuable for assisting in person identification. However, tattoo search in a large collection of unconstrained images remains a difficult problem, and existing tattoo search methods mainly focus on matching cropped tattoos, which is different from real application scenarios. To close the gap, we propose an efficient tattoo search approach that is able to learn tattoo detection and compact representation jointly in a single convolutional neural network (CNN) via multi-task learning. While the features in the backbone network are shared by both tattoo detection and compact representation learning, individual latent layers of each sub-network optimize the shared features toward the detection and feature learning tasks, respectively. We resolve the small batch size issue inside the joint tattoo detection and compact representation learning network via random image stitch and preceding feature buffering. We evaluate the proposed tattoo search system using multiple public-domain tattoo benchmarks, and a gallery set with about 300K distracter tattoo images compiled from these datasets and images from the Internet. In addition, we also introduce a tattoo sketch dataset containing 300 tattoos for sketch-based tattoo search. Experimental results show that the proposed approach has superior performance in tattoo detection and tattoo search at scale compared to several state-of-the-art tattoo retrieval algorithms.
A novel polyoxometalate-based metal organic framework (POMOF) constructed from isolated isopolyoxotungstate [H2W11O38](8-) cluster, {[Cu2(bpy)(H2O)5.5]2[H2W11O38]·3H2O·0.5CH3CN} (1, where bpy = 4,4'-bpydine), has been synthesized under solvothermal conditions and charaterized by elemental analysis, infrared spectroscopy, and single-crystal X-ray diffraction. In 1, {W11} clusters are alternately linked by two [Cu(2)(H2O)1.5(Ot)3(N)](2+) cations in an unexpected end-to-end fashion leading to a one-dimensional (1D) chain. Adjacent 1D chains are linked through Cu(1)-bpy-Cu(2) in an opposite direction to form a two-dimensional (2D) wavelike sheet along the ab plane. These 2D sheets are further stacked in a parallel fashion giving rise to the 1D channels with copper(II) cations aligned in the channels. The resulting POMOF acted as a Lewis acid catalyst through a heterogeneous manner to prompt cyanosilylation with excellent efficiency.
With the increase of the scale and complexity of massive data, data dimensionality reduction technologies, such as principal component analysis, have developed rapidly. The performance of dimension reduction technologies still needs to be further improved. In the paper we proposed a new dimensionality reduction method (Y-SPCR) based Supervised Principal Component Regression (SPCR) and Y-aware Principal Component Regression (Y-aware PCR). Experimental results on four gene expression data sets show that Y-SPCR effectively overcomes the shortcomings of SPCR and Y-aware PCR and improves the accuracy and stability of on gene expression data classification.
Rational self-assembly of hexaniobate Lindqvist-type precursor [HNb6O19]7- with soluble Cu2+ salts utilizing different strategies produces a series of giant polyniobate clusters, namely, (H2en)1.25[Cu(en)2(H2O)]2Cl4[Nb24O72H21.5]7 H2O (1; en: ethylenediamine), [Cu(en)2]3[Cu(en)2(H2O)]9[{H2Nb6O19} subset{[({KNb24O72H10.25}{Cu(en)2})2{Cu3(en)3(H2O)3}{Na1.5Cu1.5(H2O)8}{Cu(en)2}4]6}]144 H2O (2), K12Na4[H23NaO8Cu24(Nb7O22)8]106 H2O (3), and K16Na12[H9Cu25.5O8(Nb7O22)8] 73.5 H2O (4). Their structures were determined and further characterized by single-crystal X-ray diffraction analysis, IR and Raman spectroscopy, thermogravimetric analysis (TGA), and elemental analysis. Structural analyses reveal that compound 2 comprises a giant capsule anion based on a wheel-shaped cluster encapsulating a Lindqvist diprotonated cluster [H2Nb6O19]6- unit, and forms a honeycomb-like structure with the inclusion of Lindqvist-type anions [H2Nb6O19]6- in the holes, whereas 3 and 4 represent an unprecedented giant cube-shaped framework. All the compounds are built from [Nb7O22]9- fundamental building blocks. Solution Raman spectroscopy studies of 2 and 3 reveal that the solid-state structures of these polyniobate cluster anions disassemble and exist in the form of the [Nb6O19]8- unit in solution. Magnetic susceptibility measurement of 3 shows antiferromagnetic coupling interactions between CuII ions with the spin-canting phenomenon.
The methods of digital deformation observation used in seismic forecast were analyzed and appraised.The distinguishing methods of mediumshort and shortimpending anomaly were chiefly analyzed.Then these methods were appraised by means of abnormal analogy.It shows that the method of nontide analysis,oblique rate difference information,Secondary function and selfadapting valve value suit to identify shortimpending anormaly;KL linear fitting and Fourier transformation suit to identify mediumshort anomaly.
The fuzzy variable structure dynamic Bayesian network is constructed, and a statistical method based on the sample information and a learning method of sample-free Bayesian network parameters is presented; then target recognition is realized according to network inference, finally, applying the traditional hard decision, The dynamic decision is performed based on the soft decision principles and the network parameters' update online is finished based on linear weighted theory. Compared with classical static Bayesian network for target recognition, this approach resolves such issues as the sequential relationship of evidences at different time slice and the network inference of constant random variables. At the same time, the method not only improves believe of target recognition but also shortens the convergence period and effectively resolves error recognition problem caused by association. In addition, the network parameters' update online is finished.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)