International Journal of Smart Electrical Engineering
Volume:10 Issue: 3, Summer 2021

In this paper, an adaptive-neural free model scheme is proposed to control a widely-used nonlinear multivariable industrial system, a quadruple-tank process (QTP). The system consists of four tanks that are arranged in two upper and two lower formations. The main objective is defined as maintaining the level of the liquid in lower tanks via two pumps. Controlling this system is not an easy task since it has nonlinear dynamics, strong interaction between different channels, and highly interacted input and output variables. In the adaptive part of the proposed controller, the parameters and rules obtained from Lyapunov stability analysis, along with the estimation of nonlinear functions performed with the neural network, constitute the controller design steps. To highlight the controller's abilities, an additional object is defined, which is controlling the temperature of liquid of those two tanks by adding a heater to the QTP system as a modified system. Obviously, the interactions amongst the control loops are multiplied because the modified quadruple tank process (MQTP) system has four inputs and four outputs. One of the main contributions of this paper is the implementation of the closed-loop system. Regarding the importance of such a system in the industry and to test the controller practically, the closed-loop system is implemented in an industrial automation environment with the connection of Process Control System SIMATIC (PCS7) industrial software to MATLAB with Open Platform Communications (OPC) protocol. The effectiveness of the introduced scheme is verified by performing some experimental validation.

Sliding mode control is one of the most effective methods of controlling nonlinear systems with bounded uncertainty. Exponential convergence of tracking error is one of the most important problem of classic sliding mode control. One way to solve this problem is use of terminal sliding mode control. The great thing about terminal sliding mode control, is it’s robustness in face of model uncertainty and external disturbances while can guarantee tracking error converge to zero in finite time simultaneously. Usually terminal sliding mode controller, is limited by singularity at the origin and infinite control signal. This article attempts to the singularity problem in controlling underwater robots and decreasing the convergence time by defining a new sliding surface for terminal sliding mode controller. simulation results shows the efficiency of proposed controller as it effectively improves the convergence time and accuracy in under water robots with are faced by structural and environmental uncertainties.

Environmental monitoring via vehicle detecting using unmanned aerial vehicle (UAV) images is a challenging task, due to small-size, low-resolution, and large-scale variation of the objects. In this paper, a two-level ensemble deep learning (named 2EDL) based on Faster R-CNN (regional-based convolutional neural network) is introduced for multiple vehicle detection in UAV images. We use three CNN models (VGG16, ResNet50, and GoogLeNet) that have already pre-trained on huge auxiliary data as feature extraction tools, combined with five learning models (KNN, SVM, MLP, C4.5 Decision Tree, and Naïve Bayes), resulting 15 different base learners in two levels. The final class is obtained via a majority vote rule ensemble of these 15 models into five vehicle classes (car, van, truck, bus, trailer) or “no-vehicle”. Simulation results on the AU-AIR dataset of UAV images show the superiority of the proposed 2EDL technique against existing methods, in terms of the total accuracy, and FPR-FNR trade-off.

Face is a unique characteristic of the human. Detecting the state of the human face, due to its difficulty on the one hand and its many useful features on the other hand, is one of the most important issues in the image processing. In this paper, a five-layer perceptron artificial neural network (MLP) with a supervisor as a complete connection has been used to separate the different facial modes. Learning in the MLP network is done deeply with a high number of layers. The network has 4 class: anger, fear, happiness and surprise. First, the main points and areas of the face that are effective in detecting the state of the face are extracted by edge finding, and then, using the matching of the Fourier series diagram on the operational points of the face, the diagram of those points is obtained. From this diagram, a number of features in the form of three coefficients and an angular velocity are used for network training. Face database images with fixed backgrounds are used for network training. This network is first implemented with Matlab and then MLP layer multiplex is used to implement on FPGA. The results show that the proposed method can be implemented on FPGA platforms with low cost and limited resources, with appropriate output accuracy. In this paper, in addition to speed, accuracy has been tried to create an application system for communication between humans and computers.

Currently, industry and academia have shown much interest in security and privacy protection on the blockchain used in various applications. Attacks like privacy leakage and data loss make conventional methods vulnerable before emerging blockchain technology. Blockchain is a decentralized and tamper-resistant public ledger technology that guarantees security and data reliability in a peer-to-peer network. Many fields have employed blockchain, from the beginning cryptocurrency to the smart contract, social services, industry, and artificial intelligence. There are blockchain reports on vulnerabilities and security, but they lack a comprehensive survey in attacks, privacy, and security views. In this survey, we first briefly overviewed blockchain. Second, we discussed challenges and issues on the blockchain. Third, we focused on the blockchain attacks, including their cause and targeted area. We also displayed possible preventive measures in the blockchain attack. Finally, we conducted a systematic study on solutions to the blockchain security increase. In addition, this survey included blockchain privacy techniques.

In Software Defined Network (SDN), controller plane is separated from the data plane simplifying management. In these networks, data forwarding cannot be conducted just one controller. Therefore, it is needed to use multiple controllers in control plane. Since, switch-controller propagation delays and inter-controller latencies affect the performance, the problem of determining appropriate number of controllers as well as their suitable locations are two main challenges, which are known as NP-Hard. In this paper, a new clustering method based on K-means, K-Harmonics means and firefly algorithm named CPP-KKF is proposed for controller placement in SDN. Result obtained by CPP- KKF algorithm is benefitted by the advantages of all techniques. The proposed algorithm is evaluated on four topologies of TopologyZoo with different scales, that include Aarnet, Colt, Cognet, and DFN and the conducted simulations demonstrate that the proposed solution outperforms K-means, K-means++, Firefly and GSO algorithms in terms of aforementioned performance issues.