This paper proposed an adaptive explicit nonlinear model predictive control (AENMPC) technique using multiple estimation models with a convex combination framework [18] for a class of nonlinear MIMO systems. Here, the explicit solution for the control signal is obtained from an optimal performance index which can be formulated without online optimization. In this work, a closed-form control law is developed by approximating the tracking error in the receding horizon by its Taylor series expansion. The control performance of any model-based control technologies explicitly depends on the quality of the unknown system parameters hence an adaptive parameter estimator is used to estimate the system parameter online [16,17]. To ensure the boundedness of the estimated parameter within a predefined compact region, a projection based adaptive law is used [43]. Using an aerodynamic laboratory set-up, known as the twin-rotor multi-input multi-output system (TRMS), the effectiveness of the proposed control algorithm has been verified. The complete state information of the system to the proposed adaptive controller is given from an extended Kalman filter based state observer. The performance of the proposed adaptive control algorithm has been verified successfully in simulations as well as real-time experimental setup of the TRMS model and compared with an existing control approach.
相似文献In this paper, an analytical model of squeeze film damping (SQFD) of perforated plates in the free molecular regime is developed, which is based on: (1) the modification of the perforated energy transfer model (P-ETM) (Li and Hu, J Micromech Microeng 21:025006, 2011) by giving the probability of molecules entering the gap through holes; (2) the application of Sumali’s formula (J Micromech Microeng 17:2231–2240, 2007) to relate to the Monte Carlo model (MC) (Hutcherson and Ye, J Micromech Microeng 14:1726–1733, 2004) quantitatively. The analytical model can model the perforation effect on SQFD of plates of various hole sizes. Compared with experiment data and numerical models, the analytical model is proved to be accurate, easy to operate. The effect of gap distance on SQFD of perforated plate in the free molecular regime is discussed. Due to perforation effect, as gap distance increases, the damping constant of non-perforated plate decreases faster than that of perforated plate of the same size.
相似文献We consider a variational model analyzed in March and Riey (Inverse Probl Imag 11(6): 997–1025, 2017) for simultaneous video inpainting and motion estimation. The model has applications in the field of recovery of missing data in archive film materials. A gray-value video content is reconstructed in a spatiotemporal region where the video data is lost. A variational method for motion compensated video inpainting is used, which is based on the simultaneous estimation of apparent motion in the video data. Apparent motion is mathematically described by a vector field of velocity, denoted optical flow, which is estimated through gray-value variations of the video data. The functional to be minimized is defined on a space of vector valued functions of bounded variation and the relaxation method of the Calculus of Variations is used. We introduce in the functional analyzed in March and Riey(Inverse Probl Imag 11(6): 997–1025, 2017) a suitable positive weight, and we show that diagonal minimizing sequences of the functional converge, up to subsequences as the weight tends to infinity, to minimizers of an appropriate limit functional. Such a limit functional is the relaxed version of a functional, modified with suitable improvements, proposed by Lauze and Nielsen (2004) and which permits an accurate joint reconstruction both of the optical flow and of the video content.
相似文献We study the phenomenon of deliberate inter-modal interactions in a specially designed index guided Microstructured Optical Fiber (MOF) by exploiting multipole expansion method (White et al. 2002). The MOF is designed in such a way that the first layer of holes is judiciously filled with a material having refractive index slightly greater than the background material or core and remaining holes are filled with air. Accordingly, we find an interesting phenomenon of mode crossing between the fundamental mode and a targeted defect mode while tuning the wavelength. Exploring this transition wavelength of the mode crossing, we propose a design of a fiber optic sensor for refractive index measurement (Silva et al. 2014) with enhanced sensitivity.
相似文献As of July 17, 2020, more than thirteen million people have been diagnosed with the Novel Coronavirus (COVID-19), and half a million people have already lost their lives due to this infectious disease. The World Health Organization declared the COVID-19 outbreak as a pandemic on March 11, 2020. Since then, social media platforms have experienced an exponential rise in the content related to the pandemic. In the past, Twitter data have been observed to be indispensable in the extraction of situational awareness information relating to any crisis. This paper presents COV19Tweets Dataset (Lamsal 2020a), a large-scale Twitter dataset with more than 310 million COVID-19 specific English language tweets and their sentiment scores. The dataset’s geo version, the GeoCOV19Tweets Dataset (Lamsal 2020b), is also presented. The paper discusses the datasets’ design in detail, and the tweets in both the datasets are analyzed. The datasets are released publicly, anticipating that they would contribute to a better understanding of spatial and temporal dimensions of the public discourse related to the ongoing pandemic. As per the stats, the datasets (Lamsal 2020a, 2020b) have been accessed over 74.5k times, collectively.
相似文献In this work, we study the performance of state-of-the-art access methods to efficiently store and retrieve trajectories in spatial networks. First, we study how efficiently such methods can manage trajectory data to support indexing for data demanding applications where trajectory retrieval must be fast. At the same time, trajectory insertions, deletions and modifications should also be executed efficiently. Secondly, we compare the performance of progressive processing of trajectory similarity top-k queries, which is a common query in spatial applications. Specifically, we examine FNR-trees (Frentzos 2003) and MON-trees (de Almeida and Gueting, 2005), which have been proposed for trajectory management, against a novel variation of our proposed Cluster-extended Adjacency Lists (CeAL) (Tiakas and Rafailidis 2015). In particular: (a) we extend the above access methods to efficiently handle trajectories of objects that move in large spatial networks, and (b) to enhance their performance, we create an entirely new implementation framework to generate trajectories and to test the trajectory management and retrieval for each approach. With respect to the generation of trajectories, we extend the generator by Brinkhoff (2000) to efficiently support very large spatial networks. Finally, we conduct extensive experimentation which demonstrates that the proposed method CeAL prevails in space and time complexity.
相似文献The three-dimensional convolutional neural network is widely used in video recognition, action recognition and other tasks because it can directly extract temporal and spatial features. Due to the large number of parameters, many computing resources, and difficulty in training, the structure of three-dimensional convolutional neural network is generally shallow. For example, the traditional C3D [17] method uses only the 11-layer VGGNet structure, and the traditional Res3D [18] method adopts a residual network of 18 and 34 layers. Some experience of two-dimensional convolutional neural network shows that the deeper the network structure is, the higher the recognition accuracy will be. Therefore, this paper proposes a new method 3D ResNet-66, which combines a 50-layer 3D residual network and four-layer residual blocks, effectively reducing the number of parameters while increasing the depth of the network, and we finally obtain a better video recognition model through experiments. We evaluate our method on shipping event datasets. Compared to the traditional C3D and Res3D method, our method has improved the accuracy from 91.48% to 96.33%, the model size has been reduced from 561 MB to 135 MB, and the average processing time has become half of the original.
相似文献We study the problem of training deep fully connected neural networks with Rectified Linear Unit (ReLU) activation function and cross entropy loss function for binary classification using gradient descent. We show that with proper random weight initialization, gradient descent can find the global minima of the training loss for an over-parameterized deep ReLU network, under certain assumption on the training data. The key idea of our proof is that Gaussian random initialization followed by gradient descent produces a sequence of iterates that stay inside a small perturbation region centered at the initial weights, in which the training loss function of the deep ReLU networks enjoys nice local curvature properties that ensure the global convergence of gradient descent. At the core of our proof technique is (1) a milder assumption on the training data; (2) a sharp analysis of the trajectory length for gradient descent; and (3) a finer characterization of the size of the perturbation region. Compared with the concurrent work (Allen-Zhu et al. in A convergence theory for deep learning via over-parameterization, 2018a; Du et al. in Gradient descent finds global minima of deep neural networks, 2018a) along this line, our result relies on milder over-parameterization condition on the neural network width, and enjoys faster global convergence rate of gradient descent for training deep neural networks.
相似文献In this paper, we propose a mildly shock-capturing stabilized local meshless method (SLMM) for convection-dominated steady and unsteady PDEs. This work is extension of the numerical procedure, which was designed only for steady state convection-dominated PDEs (Siraj-ul-Islam and Singh in Int J Comput Methods 14(6):1750067, 2017). The proposed meshless methodology is based on employing different type of stencils embodying the already known flow direction. Numerical experiments are performed to compare the proposed method with the finite-difference method on special grid (FDSG) and other numerical methods. Numerical results confirm that the new approach is accurate and efficient for solving a wide class of one- and two- dimensional convection-dominated PDEs having sharp corners and jump discontinuities. Performance of the SLMM is found better in some cases and comparable in other cases, than the mesh-based numerical methods.
相似文献C-Mantec neural network constructive algorithm Ortega (C-Mantec neural network algorithm implementation on MATLAB. https://github.com/IvanGGomez/CmantecPaco, 2015) creates very compact architectures with generalization capabilities similar to feed-forward networks trained by the well-known back-propagation algorithm. Nevertheless, constructive algorithms suffer much from the problem of overfitting, and thus, in this work the learning procedure is first analyzed for networks created by this algorithm with the aim of trying to understand the training dynamics that will permit optimization possibilities. Secondly, several optimization strategies are analyzed for the position of class separating hyperplanes, and the results analyzed on a set of public domain benchmark data sets. The results indicate that with these modifications a small increase in prediction accuracy of C-Mantec can be obtained but in general this was not better when compared to a standard support vector machine, except in some cases when a mixed strategy is used.
相似文献While research in HCI on dealing with cultural issues when designing ICTs tended to adopt fixed and taxonomic views, recent theoretical perspectives closer to the social sciences have called for attending to the contingent, fluid, and dynamic aspects of the notion of culture. In this article, we contribute to translating these perspectives into an approach for informing design. We focus on abandoning prior conceptions of culture to allow the discovery of cultural differences through inductive field research while engaging with the target community. This allows a view on cultural difference that is generative for design: it is unique to each case, and it also remains close to the concerns of community members. We base our approach on Basile Zimmermann’s (2015) waves and forms framework, and we illustrate it through our engagement and design with VOCI, a local voluntary community of tech-savvy university students in Syria between 2011 and 2015.
相似文献Hot embossing, a polymer molding process conceived by Forschungszentrum Karlsruhe, is one of the established replication processes for microstructures The process is especially well suited for manufacturing small and medium series of microcomponents (SPIE Conference 1997; Polymer News 25:224–229, 2000; J Micromech Microeng 14:R1–14, 2004; Sensors Actuators 3:130–135, 2000). However, a wider application of the process currently is seriously hampered by the lack of adequate simulation tools for process optimization and part design. This situation is becoming more critical, as the dimension of the microstructures shrink from micron and submicron levels to the nanoscale and as productivity requirements dictate the enlargement of formats to process larger numbers of devices in parallel. Based on the current scientific work (Forschungszentrum Karlsruhe, FZKA-Bericht 7058 2003; DTIP Conference Montreux 2004; Microsystem Tech 10:432–437 2004), a German–Canadian cooperation has been started. The objective of this cooperation is to fill the gap mentioned above by developing reliable computer models and simulation tools for the hot embossing process and to incorporate these models in a user-friendly computer code. The present paper will give an overview of the activities in the project. The activities related to material characterization, especially the development of a viscoelastic material model, the characterization of friction between polymer and mold during demolding, the development of an 8-in. microstructured mold, and the fabrication of nanostructured molds will be discussed.
相似文献In this paper, a new way of computing the constraint transfer matrix for the inverse dynamic and joint force analysis of multibody systems is developed. The method is based on the Newton–Euler method and the screw theory notations. This method is first developed in (Taghvaeipour et al. in Multibody Syst. Dyn. 29(2):139–168, 2013), however, in this study, it is efficiently modified by incorporating a unified constraint transfer matrix for all types of joints. This change makes both the derivation of the equations and the computations less time consuming. Moreover, in the foregoing procedure, the constraint wrenches of a system are obtained in one reference frame, namely, the global reference frame. As a case study, the proposed method is carried out on the agile wrist which is a three-legged spherical parallel robot with three degrees of freedom. At the end, the results obtained by the modified method are verified with the ones calculated by the original procedure and a software package.
相似文献In this paper, we propose an efficient cascaded model for sign language recognition taking benefit from spatio-temporal hand-based information using deep learning approaches, especially Single Shot Detector (SSD), Convolutional Neural Network (CNN), and Long Short Term Memory (LSTM), from videos. Our simple yet efficient and accurate model includes two main parts: hand detection and sign recognition. Three types of spatial features, including hand features, Extra Spatial Hand Relation (ESHR) features, and Hand Pose (HP) features, have been fused in the model to feed to LSTM for temporal features extraction. We train SSD model for hand detection using some videos collected from five online sign dictionaries. Our model is evaluated on our proposed dataset (Rastgoo et al., Expert Syst Appl 150: 113336, 2020), including 10’000 sign videos for 100 Persian sign using 10 contributors in 10 different backgrounds, and isoGD dataset. Using the 5-fold cross-validation method, our model outperforms state-of-the-art alternatives in sign language recognition
相似文献The present article presents the hydromagnetic nanofluid flow past a stretching cylinder embedded in non-Darcian Forchheimer porous media by using Buongiorno’s mathematical model (Buongiorno in J Heat Transf 128:240–250, 2006; Nadeem et al. in J Taiwan Inst Chem Eng 45:121, 2014, Nadeem et al. Appl Nanosci 4:625–631, 2014). Thermal radiation via Roseland’s approximation (Akbar et al. in Chin J Aeronaut 26:1389–1397, 2013; Nadeem and Haq in J Aerosp Eng 28:04014061, 2012), Brownian motion, thermophoresis and Joule heating effects are also considered. To explore thermal characteristics, prescribed heat flux and prescribed mass flux boundary conditions are deployed. Governing flow problem consists of PDEs in the cylindrical form, which are converted into system of nonlinear ODEs by applying applicable similarity transforms. ODEs are tackled by RK–Fehlberg fourth–fifth-order numerical integration scheme with shooting algorithm. Impact of numerous involving physical parameters on flow features like temperature distribution, velocity distribution, Sherwood number, local Nusselt number and skin friction coefficient is shown through graphs and tables.
相似文献