Tech Science Press
Not a member yet
3972 research outputs found
Sort by
A Bio-Inspired Global Finite Time Tracking Control of Four-Rotor Test Bench System
Get PDFA bio-inspired global finite time control using global fast-terminal sliding mode controller and radial basis function network is presented in this article, to address the attitude tracking control problem of the three degree-of-freedom four-rotor hover system. The proposed controller provides convergence of system states in a pre-determined finite time and estimates the unmodeled dynamics of the four-rotor system. Dynamic model of the four-rotor system is derived with Newton’s force equations. The unknown dynamics of four-rotor systems are estimated using Radial basis function. The bio-inspired global fast terminal sliding mode controller is proposed to provide chattering free finite time error convergence and to provide optimal tracking of the attitude angles while being subjected to unknown dynamics. The global stability proof of the designed controller is provided on the basis of Lyapunov stability theorem. The proposed controller is validated by (i) conducting an experiment through implementing it on the laboratory-based hover system, and (ii) through simulations. Performance of the proposed control scheme is also compared with classical and intelligent controllers. The performance comparison exhibits that the designed controller has quick transient response and improved chattering free steady state performance. The proposed bio-inspired global fast terminal sliding mode controller offers improved estimation and better tracking performance than the traditional controllers. In addition, the proposed controller is computationally cost effective and can be implanted on multirotor unmanned air vehicles with limited computational processing capabilities
Symmetric Learning Data Augmentation Model for Underwater Target Noise Data Expansion
Get PDFAn important issue for deep learning models is the acquisition of training of data. Without abundant data from a real production environment for training, deep learning models would not be as widely used as they are today. However, the cost of obtaining abundant real-world environment is high, especially for underwater environments. It is more straightforward to simulate data that is closed to that from real environment. In this paper, a simple and easy symmetric learning data augmentation model (SLDAM) is proposed for underwater target radiate-noise data expansion and generation. The SLDAM, taking the optimal classifier of an initial dataset as the discriminator, makes use of the structure of the classifier to construct a symmetric generator based on antagonistic generation. It generates data similar to the initial dataset that can be used to supplement training data sets. This model has taken into consideration feature loss and sample loss function in model training, and is able to reduce the dependence of the generation and expansion on the feature set. We verified that the SLDAM is able to data expansion with low calculation complexity. Our results showed that the SLDAM is able to generate new data without compromising data recognition accuracy, for practical application in a production environment
A Novel Ensemble Learning Algorithm Based on D-S Evidence Theory for IoT Security
Get PDFIn the last decade, IoT has been widely used in smart cities, autonomous driving and Industry 4.0, which lead to improve efficiency, reliability, security and economic benefits. However, with the rapid development of new technologies, such as cognitive communication, cloud computing, quantum computing and big data, the IoT security is being confronted with a series of new threats and challenges. IoT device identification via Radio Frequency Fingerprinting (RFF) extracting from radio signals is a physical-layer method for IoT security. In physical-layer, RFF is a unique characteristic of IoT device themselves, which can difficultly be tampered. Just as people’s unique fingerprinting, different IoT devices exhibit different RFF which can be used for identification and authentication. In this paper, the structure of IoT device identification is proposed, the key technologies such as signal detection, RFF extraction, and classification model is discussed. Especially, based on the random forest and Dempster-Shafer evidence algorithm, a novel ensemble learning algorithm is proposed. Through theoretical modeling and experimental verification, the reliability and differentiability of RFF are extracted and verified, the classification result is shown under the real IoT device environments
Poly(phenylene oxide) and Renewable Polyamide 11 Blends Compatibilized by Ethylene-n-Octene Copolymer
No full textPoly(phenylene oxide)/renewable polyamide 11 (PPO/PA11 20/80) blends toughened with glycidyl methacrylate grafted ethylene-n-octene copolymer (GEOC) were prepared in a co-rotating twin-screw extruder. The reaction between GEOC and PPO/PA11 blend was analyzed by gel content tests. The morphology of PPO/PA11/GEOC blends was observed by scanning electron microscope. The SEM results showed that PPO formed the continuous phase, though it is a minority component of blends. With increasing GEOC content from 5 to 15 wt% the morphology of the blends transformed from droplet-matrix to co-continuous structure, in which both PA11 and PPO phases are continuous. The blend with co-continuous morphology had better mechanical properties than those with droplet-matrix morphology. The impact strength of the PPO/PA11/GEOC blends was much higher compared to the one without GEOC as well as PA11 due to the compatibilizing effect, which was also proved by DSC analysis, rheological behavior (MFR, DMTA) and tensile properties
Numerical Simulation of High Speed Rotating Waterjet Flow Field in a Semi Enclosed Vacuum Chamber
Get PDFIn this paper, a three dimension model is built according to real surface cleaner in airport runway rubber mark cleaning vehicle and numerical simulation of this model is carried out using Ansys Fluent software. After comparison and analysis the flow field between high speed rotating waterjet and static waterjet formerly studied by other researchers, the influences of different standoff distance from nozzle outlet to runway surface and rotation speed on rubber mark cleaning effect is simulated and analyzed. Results show the optimal operation parameters for the simulated model and quantitative advices are given for design, manufacture and operation of the airport runway rubber mark cleaning vehicle
A Novel Interacting Multiple-Model Method and Its Application to Moisture Content Prediction of ASP Flooding
Get PDFIn this paper, an interacting multiple-model (IMM) method based on data-driven identification model is proposed for the prediction of nonlinear dynamic systems. Firstly, two basic models are selected as combination components due to their proved effectiveness. One is Gaussian process (GP) model, which can provide the predictive variance of the predicted output and only has several optimizing parameters. The other is regularized extreme learning machine (RELM) model, which can improve the over-fitting problem resulted by empirical risk minimization principle and enhances the overall generalization performance. Then both of the models are updated continually using meaningful new data selected by data selection methods. Furthermore, recursive methods are employed in the two models to reduce the computational burden caused by continuous renewal. Finally, the two models are combined in IMM algorithm to realize the hybrid prediction, which can avoid the error accumulation in the single-model prediction. In order to verify the performance, the proposed method is applied to the prediction of moisture content of alkali-surfactant-polymer (ASP) flooding. The simulation results show that the proposed model can match the process very well. And IMM algorithm can outperform its components and provide a nice improvement in accuracy and robustness
An Ensemble Based Hand Vein Pattern Authentication System
Get PDFAmongst several biometric traits, Vein pattern biometric has drawn much attention among researchers and diverse users. It gains its importance due to its difficulty in reproduction and inherent security advantages. Many research papers have dealt with the topic of new generation biometric solutions such as iris and vein biometrics. However, most implementations have been based on small datasets due to the difficulties in obtaining samples. In this paper, a deeper study has been conducted on previously suggested methods based on Convolutional Neural Networks (CNN) using a larger dataset. Also, modifications are suggested for implementation using ensemble methods. Ensembles were used to reduce training time and cost by training multiple weak classifiers instead of a single, strong classifier. Classifiers used were CNN, Random Forest and Logistic Regression. An inexpensive and robust data acquisition system was also developed for obtaining the dataset. The obtained result shows an improved accuracy of 96.77% using ensemble method instead of dealing with a single classifier
Fluid-Structure Interaction Simulation of Aqueous Outflow System in Response to Juxtacanalicular Meshwork Permeability Changes with a Two-Way Coupled Method
Get PDFElevated intraocular pressure appears to have a broader impact on increased resistance to aqueous humor outflow through the conventional aqueous outflow system (AOS). However, there is still no consensus about exact location of the increased outflow resistance of aqueous humor, and the mechanism is not perfect. In addition, it is difficult to accurately obtain hydrodynamic parameters of aqueous humor within the trabecular meshwork outflow pathways based on the current technology. In this paper, a two-way fluid-structure interaction simulation was performed to study the pressure difference and velocity in the superficial trabecular meshwork, juxtacanalicular meshwork (JCM) and Schlemm’s canal in response to JCM permeability changes. We obtained the JCM permeability of normal intraocular pressure varied between 1×10−15 m2 and 10×10−15 m2 while permeability of the JCM ranged from 2×10−16 m2 and 3×10−16 m2 under conditions of high intraocular pressure. The study indicated that the fluid dynamics parameters in trabecular meshwork and Schlemm’s canal are most significantly affected by the changes of JCM permeability. Moreover, the study demonstrates that the finite element modeling of AOS provides a practical means for studying the outflow dynamics and the biomechanical environment of the AOS
A Numerical Study of Passive Receptor-Mediated Endocytosis of Nanoparticles: The Effect of Mechanical Properties
Get PDFIn this work, a three-dimensional axisymmetric model with nanoparticle, receptor-ligand bonds and cell membrane as a system was used to study the quasi-static receptor-mediated endocytosis process of spherical nanoparticles in drug delivery. The minimization of the system energy function was carried out numerically, and the deformations of nanoparticle, receptor-ligand bonds and cell membrane were predicted. Results show that passive endocytosis may fail due to the rupture of receptor-ligand bonds during the wrapping process, and the size and rigidity of nanoparticles affect the total deformation energy and the terminal wrapping stage. Our results suggest that, in addition to the energy requirement, the success of passive endocytosis also depends on the maximum strength of the receptor-ligand bonds
A Distributed Intrusion Detection Model via Nondestructive Partitioning and Balanced Allocation for Big Data
Get PDFThere are two key issues in distributed intrusion detection system, that is, maintaining load balance of system and protecting data integrity. To address these issues, this paper proposes a new distributed intrusion detection model for big data based on nondestructive partitioning and balanced allocation. A data allocation strategy based on capacity and workload is introduced to achieve local load balance, and a dynamic load adjustment strategy is adopted to maintain global load balance of cluster. Moreover, data integrity is protected by using session reassemble and session partitioning. The simulation results show that the new model enjoys favorable advantages such as good load balance, higher detection rate and detection efficiency