1,721,168 research outputs found
Automated Generation Control of Multiple-Area Electrical System with an Availability-Based Tariff Pricing Scheme Regulated by Whale Optimized Fuzzy PID Controller
Get PDFIn this research, a whale-optimized fuzzy PID controller was developed to manage automatic generation control in multiple-area electrical energy systems with an availability-based tariff (ABT) pricing scheme. The objective of this work is to minimize the power production costs, area control errors (ACEs), and marginal costs of the multiple-area electrical energy system with real-time load and frequency variation conditions. The generation of power, deviation of power in the tie line, and deviation of frequency of the interconnected three-area electrical energy system, including the hydrothermal steam power plant and gas power plant, will be measured and analyzed rigorously. Based on the output from the whale optimization, the fuzzy PID controller regulates the deviation of power in the tie line and the deviation of frequency of the interconnected three-area electrical energy system. The reliability and suitability of the proposed optimization, i.e., whale-optimized fuzzy PID controller, are investigated against already presented methods such as particle swarm optimization and genetic algorithms
Performance analysis the effect of network size on reactive and geographical manet routing protocols
No full textSince the nodes inside the network may be quickly deployed and modified as necessary, the
Mobile Ad hoc Network (MANET) is especially useful in distant places with limited access
to conventional wired networks. A self-configuring net thought up from mobile routers and
related hosts that are linked through wireless links is known as MANET. MANET is a kind
of mobile ad-hoc network. There are many routing protocols that are being implemented
and studied for MANET; these protocols represent one of the main challenges for MANET.
In this thesis the influence of MANET on net dimension routing protocols will be researched
and assessed. With varied circumstances, we applied different methods. The reactive group
tried ad hoc on-demand remote vector routing as a routing protocol (AODV). Also, this
avaricious stateless routing perimeter is popular among the geographical group (GPSR). To
evaluate each protocol, we employed four performance indicators (PDR, throughput, E2ED,
and NRO). The NS-2 simulator was utilized in the creation of our recommended model. In
the findings, we determined that AODV is the best protocol for E2ED while GPSR is the
best for throughput, PDF and NRO. The results lead us to the conclusion that each approach
is appropriate for a certain setting. In the findings, we determined that AODV is the best
protocol for E2ED while GPSR is the best for throughput, PDF and NRO. The results lead
us to the conclusion that each approach is appropriate for a certain setting
Age and gender classification in EEG signals using deep learning
No full textThe most recent advancements in this field have made it possible for a vast array of BCIbased applications to be created, which has led to an explosion in the number of such
applications. In this study, we suggest a novel application of BCI, which entails estimating
a person's age and gender based on the analysis of their electroencephalogram (EEG). The
brainwave activity of sixty distinct individuals, both male and female, was recorded while
they were sleeping peacefully with their eyes closed by using equipment that is considered
to be standard EEG recording gear. A hybrid learning architecture was constructed with the
aid of a deep CNN network so that this analysis could be carried out
Modeling analysis and control of mixed source in the microgrid environment to development of ensemble machine learning platform for the energy management system
Get PDFImprovement of the power grid (PG) for energy system and solar charging is rectifiers remains
challenging. We propose an optimal design of smart microgrid for efficient functioning of the
energy management system. This design is based on a synergistic combination of the machine
learning with Neural Network and Robust control schemes. The best parameters of power grid
and robust control are determined via optimization, where NN is tuned using genetic algorithm
to achieve the optimal solution. NN is used to enhance the robust control parameters for
designing NN of the Machine learning system. The entire scheme is further tuned by hybrid
energy parameters under various operating conditions to improve the power grid management
performance in terms of charging and rectifying. Performing the proposed analog-implemented
energy management controller is evaluated by interfacing it with a hardware prototype
experimental application of dual photovoltaic (PV) system
Efficient routing in VANET network using metuheuristic based modification of maodv protocol
Get PDFVANET (Vehicular Ad hoc Network) is a new technology that uses vehicles as nodes to create
a mobile network. These vehicles are equipped with wireless interfaces allowing them to
communicate with each other. Indeed, VANETs can be used to extend the scope of safety
information (alert messages, information on anomalies, etc.) or other types of applications
(multimedia, etc.). On the basis of previous remark, we propose in this thesis solutions based on
metaheuristics. heuristic algorithms make it possible to find feasible solutions having an
acceptable spatial and temporal complexity. It is a technique to improve the time to determine
an optimal solution for NP-difficile problems. First, we analyze the applicable heuristics for NPhard problems. We detail the performance and characteristics of each. Next, we use the most
appropriate for the vehicular routing problem
The combination of cyber security, excessive programming, and available technology to create an effective weapon
Get PDFThis research presents a model for designing a military terrestrial drone from available
technology by using raspberry pi, the proposed model should fulfil the general requirements
by providing a secure remote connection, a camera control application with a reliable
targeting system, and motion control. We suggested SSH connection with cyber-attack
simulation by Ettercap and sslstrip tool to test its reliability, and mark vulnerabilities if exist,
as for the camera application and targeting system, python was our main language to write
it and we used SSD object detection method of machine learning for camera to be able to
identify and track objects it sees, python libraries like, Tkinter along with PIL library was
used to create the application GUI interface and its functions along with OpenCV to provide
algorithms for video stream processing and displaying, we used OpenCV also for object
detection algorithms writing, and socket was used to create client-server connection on
raspberry pi and windows controlling machine to transfer the targeting control signal, as for
the controlling of motion functions its done through SSH PuTTY app and the python curses
library. The research’s goal is to prove a theory regarding the military technology in the
middle east of achieving great efficiency by mostly using available and civilian-accessed
technology
Enhancing double authentication data security by face mask detection and recognition using k-means algorithm and CNN classification
Get PDFFace detection is an interesting topic to investigate since it is both practical and challenging. Face
recognition is so essential to a broad range of practical applications that solving the issue of face
detection is often the first step. Prior research efforts concentrated mostly on the challenge of
face recognition; however, face detection has now received greater consideration. This is because
face recognition systems need face detection systems, particularly for images with a chaotic
backdrop the fundamental contribution of this investigation is to categorize face characteristics
using statistical learning methods. This includes identifying and labelling as many facial traits as
possible from a series of sample images without previous knowledge of the researched qualities.
Excellent results were obtained in the identification of mouths, with detection rates above 95%
percent and mistakes equivalent to those associated with manually identifying mouths. As part of
the current study, detectors and classifiers of the most crucial properties were created. In
addition, the lens sorter performed admirably, achieving identification rates of about 93% in
databases with controlled circumstances and 90% in databases with uncontrolled settings. After
using the mouth detector, classifiers for moustaches and beards performed remarkably well.
Even with unstructured datasets, they obtained a detection rate of above 93%
Network anomaly detection in cloud and IoT using fuzzy logic
Get PDFThe Internet of Things (IoT) is a new technology that seeks to give all physical objects a
virtual presence on the Internet. The primary idea behind IoT is to give physical objects
intelligence by incorporating electronic components, software, sensors, actuators, and
network connectivity. Hence, the Internet of Things can be characterized as a network of
these smart items that can collect and share data via the Internet. As a result, this thesis
proposes a solution capable of detecting a wide range of intrusions on connected objects,
employing powerful tracing techniques to collect precise information on the monitored
systems and comparing various machine learning (ML) techniques to achieve superior
detection capabilities
Efficient solar tracking based on maximum power point using firefly algorithm and FPGA controller
Get PDFThe management system for "demand response" is included into the EMS as a standard
feature. This system use traffic lights that are installed inside of homes to communicate with
customers and tell them whether they need to reduce, raise, or keep the same level of usage in
order to achieve optimal delivery. In this research, two factors relating to this system are
explored: the calculation of the demand range, and the influence of demand control signals on
the forecast. Both of these topics are discussed in regard to this system. In order to determine
the dynamic range for the first subject based on the data from the preceding topics, firefly
intervals will be applied. The load displacement factor of the optimizer has upper and lower
constraints that are defined within this range. These boundaries are the basis for the signals
that are sent to the consumers in this study, an FPGA board with the MPPT FA algorithm is
used. The system output consists of a pulse width modulated (PWM) pulse generation
controller which is fed to a Buck type DC/DC converter circuit. This converter is the intermediary between the panel and the load, responsible for voltage and current adequacy and
power transfer
Wind turbine energy system control with particle swarm optimization
Get PDFNowadays, the use of green energy is now more important in the world. Wind power is one of
these green renewable energies and is a serious competitor of fossil fuel for electricity market.
Also, the cost of production should be comparable to fossil fuels and other fuelsin this competition.
Infrastructure and machinery are of first importance for capital investment. For example, turbine
design is an important factor that reduces energy costs, so as it can be understood, construction
and operation are important factors that make wind energy competitive. The air flow of the wind
and its mathematics are significant for understanding operating efficiency of wind turbine in a
specific area because it leads to optimize and help for enhancing these control algorithms for
inspection. The modeling has affected the change in the performance of the wind turbine. This
thesis aims to model the wind turbine by examining the general information and these parameters
in the light of the power coefficient factor. The coherent model results obtained from here aim to
reduce production costs and become economical by helping designers and new generationturbine
researchers to design and optimize these turbines
- …
