Target detection and search requires a target recognition algorithm and a target search algorithm on a multimedia wireless sensor network (MWSN). One of the main problems in MWSN is energy efficiency of image transmis...
详细信息
In the future, active network management (ANM) and adaptive distributed energy resources (DER) control schemes will be increasingly required to fulfill growing power system resiliency and renewable generation integrat...
详细信息
ISBN:
(纸本)9781728176604
In the future, active network management (ANM) and adaptive distributed energy resources (DER) control schemes will be increasingly required to fulfill growing power system resiliency and renewable generation integration needs. As part of future ANM schemes more frequent electricity distribution network topology changes could be utilized to improve the electricity supply reliability. However, these topology changes may require DER control methods and functions adaptation accordingly in order to maintain feasible power quality, voltage level and supply reliability in the distribution network. this paper presents a real-life case study from the needed adaptation of DER control schemes and functions after MV distribution network topology change from normal to back-up connection. the focus in the studied case is on MV network connected wind turbine (WT) control scheme adaptation after the topology change based on real measured data from the WT. In addition, the paper presents also simulations from potential future scenarios with centralized and distributed battery energy storage systems (BESSs) which could be also simultaneously utilized for the provision of local and system-wide services.
Data-driven fault classification for induction machines has received much attention in electric drives. In this study, a data-driven and supervised machine learning-based fault classification technique is addressed by...
详细信息
ISBN:
(纸本)9781665427302
Data-driven fault classification for induction machines has received much attention in electric drives. In this study, a data-driven and supervised machine learning-based fault classification technique is addressed by integrating t-distributed stochastic neighbour embedding (t-SNE) and support vector machine (SVM) to evaluate the feasibility and capability of the classification performances. the algorithm proposed is applied to the three-phase induction machine control systems subjected to stator inter-turn faults, including single phase and multi-phase faults with different values of fault ratios. Finally, intensive simulations and comparison studies are presented to validate the classification method.
the political and ideological course resource recommendation is critical in enhancing the efficiency of university and college education. Nonetheless, existing recommendation mechanisms, including collaborative filter...
详细信息
ISBN:
(数字)9798331533663
ISBN:
(纸本)9798331533670
the political and ideological course resource recommendation is critical in enhancing the efficiency of university and college education. Nonetheless, existing recommendation mechanisms, including collaborative filtering and content-based methods, are plagued by issues such as data sparsity, cold-start issues, and lack of adaptability to the changing learning interests of students. these problems result in low-quality and less personalized recommendations, decreasing student participation and learning achievements. To overcome these difficulties, this research recommends a deep reinforcement learning-based recommendation model, employing the Q-learning algorithm. In contrast to traditional methods, the recommended model learns from students' past interactions and feedback constantly to dynamically adjust course recommendations. By applying a reward-based mechanism, the system guarantees adaptive learning activities and assigns students with suitable ideological and political course materials that align withtheir interests and academic requirements. Massive experiments on actual educational datasets prove that the designed model far exceeds conventional recommendation methods in precision, relevance, and user satisfaction. Experimental results indicate enhanced recommendation accuracy Experimental results indicate enhanced recommendation accuracy of 0.993, precision of 0.989, recall of 0.978 and F1-score of 0.983, ultimately creating a better learning environment. this work makes a contribution to the field of intelligent education by promoting greater personalization of ideological and political course suggestions. Hybrid reinforcement learning strategies and multi-agent cooperation can be further investigated in future research to enhance recommendation effectiveness and scalability.
Internet of things (IoT) devices are the weak link in organizing a Wireless sensor Network. Various Attacks on IoT devices can lead to different complex consequences. Real applications of the IoT generate a large amou...
详细信息
ISBN:
(数字)9798350374865
ISBN:
(纸本)9798350374872
Internet of things (IoT) devices are the weak link in organizing a Wireless sensor Network. Various Attacks on IoT devices can lead to different complex consequences. Real applications of the IoT generate a large amount of data every second, the confidentiality of which is of very high value. therefore, detecting attacks in IoT interactions is of paramount interest from both science and industry. Among various attack detection approaches, machine learning methods show great potential due to their early detection ability. the paper presents a methodology for detecting attacks based on two machine learning methods: Random Forest (RF) and Artificial Neural Network (ANN). To conduct the experiment, various data sets were considered, the descriptions of which are given in the work. the experiment was carried out using open data sets obtained from real IoT devices. As a result, RF demonstrated a high accuracy of 98.6%.
Equivalent circuit models for fuel cell stacks as well as for the series connection of battery cells are characterized by multiple series-connected RC sub-networks. In practical applications, for example, when aiming ...
Equivalent circuit models for fuel cell stacks as well as for the series connection of battery cells are characterized by multiple series-connected RC sub-networks. In practical applications, for example, when aiming at the aging detection and monitoring of the individual cells in a fuel cell stack, it is desired to estimate the individual state variables of each of these series-connected cells. the same holds true for battery management systemsthat are dedicated to the state of charge (resp., voltage) equalization in a series connection of multiple battery cells. However, the pure knowledge of the current through this series connection as well as the sum over all terminal voltages is insufficient to make the overall system model fully observable. To efficiently observe the state variables in each of the series-connected subsystems, we aim at avoiding the measurement of the voltages of each individual fuel (resp., battery) cell. Instead, this paper presents a systematic sensor switching strategy which allows for estimating the individual state variables of all series-connected subsystems in an Unscented Kalman Filter framework. Simulation results are presented to compare the achievable estimation accuracy with a scenario in which all cell (resp., terminal) voltages were measured simultaneously.
this research focuses on the development of a measurement and control system for the main parameters of agricultural greenhouses, based on the PIC16F877 microcontroller. Installed at the University of Oran 1, the syst...
详细信息
ISBN:
(数字)9798331519872
ISBN:
(纸本)9798331519889
this research focuses on the development of a measurement and control system for the main parameters of agricultural greenhouses, based on the PIC16F877 microcontroller. Installed at the University of Oran 1, the system utilizes temperature, humidity, and light sensors to monitor the greenhouse environment. A measurement and control system is already installed, and we have developed a low cost and reliable redundancy for it. the results obtained are accurate with an acceptable tolerance of 1.2% for temperature and 2.4% for humidity. Relays have been implemented to control the direction and speed of the fans, ensuring optimal conditions within the greenhouse. After validating the board through several tests and calibrations, a commercial PCB version was produced using open source EasyEDA software for the design. To enhance data dissemination and contribute further, a WiFi module with a protocol for communication will be added, introducing a smart agricultural greenhouse.
this paper proposes a distributed passive localization system that utilizes dual-layer data fusion through smart contracts to enhance the accuracy and security of positioning in the presence of Byzantine nodes. Firstl...
详细信息
ISBN:
(数字)9798331506155
ISBN:
(纸本)9798331506162
this paper proposes a distributed passive localization system that utilizes dual-layer data fusion through smart contracts to enhance the accuracy and security of positioning in the presence of Byzantine nodes. Firstly, the system groups sensors based on distance and utilizes passive sensors for Angle of Arrival (AOA) measurements, ensuring both precise localization and operational concealment. Communication between nodes is based on blockchain technology, with encryption algorithms and hash signatures safeguarding data security. Subsequently, the proposed method sequentially carries out intra-group and inter-group data fusion, effectively detecting and mitigating the influence of Byzantine nodes, leading to highly accurate position estimations. Finally, the Practical Byzantine Fault Tolerance (PBFT) algorithm ensures consensus on the positioning results across all nodes. Experimental validation and analysis demonstrate that the proposed approach significantly enhances the accuracy, robustness, and security of distributed localization in adversarial environments, offering a reliable solution for applications requiring high levels of trust and precision.
Locational detection of the false data injection attack (FDIA) is essential for smart grid cyber-security. However, the FDIA detection techniques often falter in scalability as power network complexity increases. To a...
Locational detection of the false data injection attack (FDIA) is essential for smart grid cyber-security. However, the FDIA detection techniques often falter in scalability as power network complexity increases. To address the research gap, this paper introduces an innovative distributed framework for locational FDIA detection that optimizes both performance and scalability. the proposed framework initially partitions the power grid using the improved Louvain community detection algorithm. the proposed solution utilizes the Electrical Functional Strength (EFS) matrix and power supply modularity. Subsequently, a dedicated multi-label one-dimensional convolutional neural network model (1D CNN) locational detector is designed for each derived cluster. the proposed methodology is designed to increase detection accuracy and enhance the scalability of the model. this is achieved by reducing training and detection times, as well as lowering memory requirements, compared to traditional centralized approaches. the effectiveness of the proposed framework is validated through simulations on the ieee 39-bus system. these simulations demonstrate the framework’s capability to enhance detection accuracy by simplifying the locational FDIA detection challenge, achieved through strategic grid partitioning.
State estimation refers to the integration of sensor input and observation information to estimate the current state of a robot, such as its position and orientation. However, achieving robust and accurate state estim...
详细信息
ISBN:
(数字)9798331531225
ISBN:
(纸本)9798331531232
State estimation refers to the integration of sensor input and observation information to estimate the current state of a robot, such as its position and orientation. However, achieving robust and accurate state estimation in challenging scenarios remains an urgent problem to solve. In this paper, we propose a novel Event-Visual-Inertial Odometry (EVIO) method that tightly couples event data, standard images, and inertial measurements. this method fully leverages the excellent complementary characteristics of multi-sensor fusion to estimate the ego-motion of sensors in challenging scenarios such as High Dynamic Range (HDR), low light, and motion blur caused by vigorous motion. In challenging benchmark experiments, our method outperforms state-of-the-art methods by 16%, and this achievement does not rely on extensive manual parameter tuning. Lastly, we qualitatively demonstrate the accuracy and robustness of our method for real-time state estimation across different scales and types of challenging scenarios.
暂无评论