Downloads & Free Reading Options - Results

Radar signal processing with multilayered perceptrons was investigated. Networks with no hidden layer and a single hidden layer were tested on field collected millimeter wave target returns that have been corrupted with artificial Gaussian noise at a signal to noise level of 3 dB. Performance as a function of network architecture was characterized. Keywords: Radar signal processing, Multilayered perceptrons, Single hidden layer, Field collected millimeter wave target.

Radar signal processing with multilayered perceptrons was investigated. Networks with no hidden layer and a single hidden layer were tested on field collected millimeter wave target returns that have been corrupted with artificial Gaussian noise at a signal to noise level of 3 dB. Performance as a function of network architecture was characterized. Keywords: Radar signal processing, Multilayered perceptrons, Single hidden layer, Field collected millimeter wave target.

This report shows the feasibility of using artificial neural systems (ANS) for making predictions of cloud to ground lightning strikes. ANS designs offer some potentially useful features. ANS predictors can be incrementally trained for new levels of performance without starting programming from 'scratch' each time the predictor is upgraded. Incremental training could proceed in the field reducing costs and delays of modifications while improving predictor accuracy by tailoring it to site conditions (i.e. topography, etc). Trained ANs provides a ready-made formula for constructing fast parallel, distributed processors. The features built up within the ANS might be analyzed for clues to the physical processes underlying the partially understood phenomenon of lightning. Comparisons are made of the performance of an ANS predictor with the state-of-the-art lightning prediction using a wind convergence based criterion described by Watson et al, 1987. (jes)

This paper presents an overview and analysis of learning in Artificial Neural Systems (ANS's). It begins with a general introduction to neural networks and connectionist approaches to information processing. The basis for learning in ANS's is then described, and compared with classical Machine learning. While similar in some ways, ANS learning deviates from tradition in its dependence on the modification of individual weights to bring about changes in a knowledge representation distributed across connections in a network. This unique form of learning is analyzed from two aspects: the selection of an appropriate network architecture for representing the problem, and the choice of a suitable learning rule capable of reproducing the desired function within the given network. The various network architectures are classified, and then identified with explicit restrictions on the types of functions they are capable of representing. The learning rules, i.e., algorithms that specify how the network weights are modified, are similarly taxonomized, and where possible, the limitations inherent to specific classes of rules are outlined.

Includes bibliographical references and index

The research goals of this project were to ascertain the applicability of Artificial Neural Systems (ANS) technology to expert systems tasks in general and to support the simulation of Air Combat Maneuvering (ACM) decision-making in the training environment. In the experiments conducted under this program, neural networks have aptly displayed their unique capabilities to overcome some of the more difficult aspects of knowledge engineering. ANS approaches have been shown to be capable of producing robust, generalized solutions even under novel circumstances. By capturing and simulating the expertise of human pilots in a neural network, students may be provided with expert training devices which may come very close to the look and feel of real air-to-air combat. It is expected that ANS technology will continue to provide new solutions to the simulation of human performance for training purposes.

This paper investigates the fault tolerance characteristics of time continuous recurrent artificial neural networks (ANN) that can be used to solve optimization problems. The principle of operations and performance of these networks are first illustrated by using well-known model problems like the traveling salesman problem and the assignment problem. The ANNs are then subjected to 13 simultaneous 'stuck at 1' or 'stuck at 0' faults for network sizes of up to 900 'neurons'. The effects of these faults is demonstrated and the cause for the observed fault tolerance is discussed. An application is presented in which a network performs a critical task for a real-time distributed processing system by generating new task allocations during the reconfiguration of the system. The performance degradation of the ANN under the presence of faults is investigated by large-scale simulations, and the potential benefits of delegating a critical task to a fault tolerant network are discussed.

Dissertation supervisor, Daniel J. Collins

The paper develops important fundamental steps in applying artificial neural networks in the design of intelligent control systems. Different architectures including single layered and multi layered of neural networks are examined for controls applications. The importance of different learning algorithms for both linear and nonlinear neural networks is discussed. The problem of generalization of the neural networks in control systems together with some possible solutions are also included.

Attempts to identify physiological control systems using traditional engineering or statistical approaches generally fails to produce generalized models. This report presents an alternative approach using a hybrid model. An artificial neural network is used to model the control system and a lumped parameter model is used to model the passive system. Two artificial (model) water bath systems and two robot arm models were developed to test the feasibility of this approach. Observed data were generated using these model systems by recording responses to simulated perturbations. The hybrid model was then fit to the observed data by adjusting neural network connection weights to minimize the error between observed and predicted values for one or more system variables. The fitting procedure was repeated three times under each set of conditions. Only 2-3 hidden layer nodes were required to simulate the artificial model systems. It was not necessary to include the control system output in the error term. The methodology was robust; successful control system identification was performed when errors (+/- 10% variance) were introduced into the observed data set and also when errors (+/- 10% variance) were introduced into the passive system parameter values.

This article is from Sensors (Basel, Switzerland) , volume 12 . Abstract In a wireless communication system, wireless location is the technique used to estimate the location of a mobile station (MS). To enhance the accuracy of MS location prediction, we propose a novel algorithm that utilizes time of arrival (TOA) measurements and the angle of arrival (AOA) information to locate MS when three base stations (BSs) are available. Artificial neural networks (ANN) are widely used techniques in various areas to overcome the problem of exclusive and nonlinear relationships. When the MS is heard by only three BSs, the proposed algorithm utilizes the intersections of three TOA circles (and the AOA line), based on various neural networks, to estimate the MS location in non-line-of-sight (NLOS) environments. Simulations were conducted to evaluate the performance of the algorithm for different NLOS error distributions. The numerical analysis and simulation results show that the proposed algorithms can obtain more precise location estimation under different NLOS environments.

This article is from Sensors (Basel, Switzerland) , volume 12 . Abstract In a wireless communication system, wireless location is the technique used to estimate the location of a mobile station (MS). To enhance the accuracy of MS location prediction, we propose a novel algorithm that utilizes time of arrival (TOA) measurements and the angle of arrival (AOA) information to locate MS when three base stations (BSs) are available. Artificial neural networks (ANN) are widely used techniques in various areas to overcome the problem of exclusive and nonlinear relationships. When the MS is heard by only three BSs, the proposed algorithm utilizes the intersections of three TOA circles (and the AOA line), based on various neural networks, to estimate the MS location in non-line-of-sight (NLOS) environments. Simulations were conducted to evaluate the performance of the algorithm for different NLOS error distributions. The numerical analysis and simulation results show that the proposed algorithms can obtain more precise location estimation under different NLOS environments.

Click here to view the University of Florida catalog record

With an increase of non-linear load in today’s electrical power systems, the rate of power quality drops and the voltage source and frequency deteriorate if not properly compensated with an appropriate device. Filters are most common techniques that employed to overcome this problem and improving power quality. In this paper an improved optimization technique of filter applies to the power system is based on a particle swarm optimization with using artificial neural network technique applied to the unified power flow quality conditioner (PSO-ANN UPQC). Design particle swarm optimization and artificial neural network together result in a very high performance of flexible AC transmission lines (FACTs) controller and it implements to the system to compensate all types of power quality disturbances. This technique is very powerful for minimization of total harmonic distortion of source voltages and currents as a limit permitted by IEEE-519. The work creates a power system model in MATLAB/Simulink program to investigate our proposed optimization technique for improving control circuit of filters. The work also has measured all power quality disturbances of the electrical arc furnace of steel factory and suggests this technique of filter to improve the power quality. 

Model predictive control (MPC) is becoming more and more popular in power electronics applications, yet its practical implementation faces challenges due to computational complexity and resource demands. To address these issues, a novel MPC control approach using an artificial neural network (ANN-MPC) is put forth in this research. Using a real-time circuit modeling environment, a power converter with a virtual MPC controller that can regulate both linear and nonlinear loads is first created and run. The input-output data gathered from the virtual MPC is then used to train an artificial neural network (ANN) offline, enabling a simplified mathematical representation that significantly reduces computational complexity. Moreover, the ANN-MPC controller’s adaptability to input variations enhances robustness against system uncertainties. We offer a thorough explanation of the ANN-MPC's fundamental idea, ANN architecture, offline training approach, and online functioning. The suggested controller is validated by simulation with MATLAB/Simulink tools. Performance evaluation of the novel MPC-ANN controller is performed across various scenarios, including linear and nonlinear loads under various operational conditions, and a comparative analysis with conventional MPC is presented. 

Recurrent and feedforward artificial neural networks are developed as wavefront reconstructors. The recurrent neural network studied is the Hopfield neural network and the feedforward neural network studied is the single layer perceptron artificial neural network. The recurrent artificial neural network input features are the wavefront sensor slope outputs and neighboring actuator feedback commands. The feedforward artificial neural network input features are just the wavefront sensor slope outputs. Both artificial neural networks use their inputs to calculate deformable mirror actuator commands. The effects of training are examined.

Recurrent and feedforward artificial neural networks are developed as wavefront reconstructors. The recurrent neural network studied is the Hopfield neural network and the feedforward neural network studied is the single layer perceptron artificial neural network. The recurrent artificial neural network input features are the wavefront sensor slope outputs and neighboring actuator feedback commands. The feedforward artificial neural network input features are just the wavefront sensor slope outputs. Both artificial neural networks use their inputs to calculate deformable mirror actuator commands. The effects of training are examined.

Over the past three years we have developed a novel neural-endocrine system capable of being deployed on actual sailing robots. The initial project plan was to develop a combined neural-immune-endocrine system, but after staffing issues in the first year of the project, work was refocussed to focus on the neural endocrine aspects. Whilst the primary focus was developing a system capable of sailing deployment, along the way, we have developed and investigated the application of neural-endocrine on land-based robotic units as well as land based swarm systems. This allowed to us to continue the work on the neural-endocrine architecture whilst at the same time developing the sailing robotic system (as there was a large amount of time consumed with the more engineering side of work on the sailing robotic system). We feel this project has been a success and we list the highlights from the project.

This article is from PLoS ONE , volume 8 . Abstract Artificial neural networks (ANNs) are processors that are trained to perform particular tasks. We couple a computational ANN with a simulated affective system in order to explore the interaction between the two. In particular, we design a simple affective system that adjusts the threshold values in the neurons of our ANN. The aim of this paper is to demonstrate that this simple affective system can control the firing rate of the ensemble of neurons in the ANN, as well as to explore the coupling between the affective system and the processes of long term potentiation (LTP) and long term depression (LTD), and the effect of the parameters of the affective system on its performance. We apply our networks with affective systems to a simple pole balancing example and briefly discuss the effect of affective systems on network performance.

This article is from PLoS ONE , volume 8 . Abstract Artificial neural networks (ANNs) are processors that are trained to perform particular tasks. We couple a computational ANN with a simulated affective system in order to explore the interaction between the two. In particular, we design a simple affective system that adjusts the threshold values in the neurons of our ANN. The aim of this paper is to demonstrate that this simple affective system can control the firing rate of the ensemble of neurons in the ANN, as well as to explore the coupling between the affective system and the processes of long term potentiation (LTP) and long term depression (LTD), and the effect of the parameters of the affective system on its performance. We apply our networks with affective systems to a simple pole balancing example and briefly discuss the effect of affective systems on network performance.

This article is from PLoS ONE , volume 8 . Abstract Artificial neural networks (ANNs) are processors that are trained to perform particular tasks. We couple a computational ANN with a simulated affective system in order to explore the interaction between the two. In particular, we design a simple affective system that adjusts the threshold values in the neurons of our ANN. The aim of this paper is to demonstrate that this simple affective system can control the firing rate of the ensemble of neurons in the ANN, as well as to explore the coupling between the affective system and the processes of long term potentiation (LTP) and long term depression (LTD), and the effect of the parameters of the affective system on its performance. We apply our networks with affective systems to a simple pole balancing example and briefly discuss the effect of affective systems on network performance.

In order to sustain monetary growth, the introduction of renewable energy into the electricity grid is crucial especially for many foreign African locations. Thus, in order to reap sustainable strength, these foreign locations may also outfit their airport electrical equipment with advanced synthetic intelligence technologies. For a distributive hybrid solar strength grid, the paper attempts to propose an actual time energy management algorithm. It continues with the combining of photovoltaic and wind energy for network simulation. As a function of the number of wind aero mills and photovoltaic solar panels, a multi goal approach is proposed to optimize the spectral efficiency of the location and the energy performance. For the MATLAB software simulation, radio criteria for Cell Wireless Interoperability Medium Access WiMAX technologies are taken into account. The obtained effects are much mitigated however theoretically encouraging for the mixing of green energy integration into the modern telecommunication structures. Shreya Deshmukh | Mr. Ashish Chourey | Dr. Ritu Shrivastava | Dr. Rajiv Srivastava "Analysis and Implementation of the Artificial Neural Network (ANN) Approach for the Integration of Solar and Wind Energy Sources into Telecommunication Systems" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-6 | Issue-5 , August 2022, URL: https://www.ijtsrd.com/papers/ijtsrd51786.pdf Paper URL: https://www.ijtsrd.com/engineering/electronics-and-communication-engineering/51786/analysis-and-implementation-of-the-artificial-neural-network-ann-approach-for-the-integration-of-solar-and-wind-energy-sources-into-telecommunication-systems/shreya-deshmukh

Massive information has been transmitted through complicated network connections around the world. Thus, providing a protected information system has fully consideration of many private and governmental institutes to prevent the attackers. The attackers block the users to access a particular network service by sending a large amount of fake traffics. Therefore, this article demonstrates two-classification models for accurate intrusion detection system (IDS). The first model develops the artificial neural network (ANN) of multilayer perceptron (MLP) with one hidden layer (MLP1) based on distributed denial of service (DDoS). The MLP1 has 38 input nodes, 11 hidden nodes, and 5 output nodes. The training of the MLP1 model is implemented with NSL-KDD dataset that has 38 features and five types of requests. The MLP1 achieves detection accuracy of 95.6%. The second model MLP2 has two hidden layers. The improved MLP2 model with the same setup achieves an accuracy of 2.2% higher than the MLP1 model. The study shows that the MLP2 model provides high classification accuracy of different request types.

Massive information has been transmitted through complicated network connections around the world. Thus, providing a protected information system has fully consideration of many private and governmental institutes to prevent the attackers. The attackers block the users to access a particular network service by sending a large amount of fake traffics. Therefore, this article demonstrates two-classification models for accurate intrusion detection system (IDS). The first model develops the artificial neural network (ANN) of multilayer perceptron (MLP) with one hidden layer (MLP1) based on distributed denial of service (DDoS). The MLP1 has 38 input nodes, 11 hidden nodes, and 5 output nodes. The training of the MLP1 model is implemented with NSL-KDD dataset that has 38 features and five types of requests. The MLP1 achieves detection accuracy of 95.6%. The second model MLP2 has two hidden layers. The improved MLP2 model with the same setup achieves an accuracy of 2.2% higher than the MLP1 model. The study shows that the MLP2 model provides high classification accuracy of different request types.

Identification and control of non-linear dynamical systems is a very complex task which requires new methods of approaching. This research addresses the problem of emulation and control via the use of distributed parallel processing, namely artificial neural networks. Four models for describing non- linear MIMO dynamical system are presented. Based on these models a combined feedforward and recurrent neural networks are structured to emulate the dynamical system. Further, a procedure to emulate multiple systems in a single network is suggested. A method for finding a minimal realization of a network is introduced. The minimization greatly reduces the complexity of the network without degrading the operating performance of the network. This work also examines the application of artificial neural networks for adaptive control. The multiple-system approach is used to find an adaptive neural network controller for non-linear MIMO time-varying system in a direct model reference control scheme. The controller network is trained using a procedure called bock- propagation through the plant, which was extended in this work. The application of neural networks is demonstrated on a longitudinal model of the F/A-18A fighter aircraft both with the undamaged aircraft and with a the mechanism as a time-varying MIMO dynamical system. Neural networks, Identification, Adaptive control, Non-Linear systems, Tim-Varying dynamical systems.

Identification and control of non-linear dynamical systems is a very complex task which requires new methods of approaching. This research addresses the problem of emulation and control via the use of distributed parallel processing, namely artificial neural networks. Four models for describing non- linear MIMO dynamical system are presented. Based on these models a combined feedforward and recurrent neural networks are structured to emulate the dynamical system. Further, a procedure to emulate multiple systems in a single network is suggested. A method for finding a minimal realization of a network is introduced. The minimization greatly reduces the complexity of the network without degrading the operating performance of the network. This work also examines the application of artificial neural networks for adaptive control. The multiple-system approach is used to find an adaptive neural network controller for non-linear MIMO time-varying system in a direct model reference control scheme. The controller network is trained using a procedure called bock- propagation through the plant, which was extended in this work. The application of neural networks is demonstrated on a longitudinal model of the F/A-18A fighter aircraft both with the undamaged aircraft and with a the mechanism as a time-varying MIMO dynamical system. Neural networks, Identification, Adaptive control, Non-Linear systems, Tim-Varying dynamical systems.

This report investigates the application of artificial neural networks to the problem of power system stability. The field of artificial intelligence, expert systems, and neural networks is reviewed. Power system operation is discussed with emphasis on stability considerations. Real-time system control has only recently been considered as applicable to stability, using conventional control methods. The report considers the use of artificial neural networks to improve the stability of the power system. The networks are considered as adjuncts and as replacements for existing controllers. The optimal kind of network to use as an adjunct to a generator exciter is discussed.

Identification and control of non-linear dynamical systems is a very complex task which requires new methods of approaching. This research addresses the problem of emulation and control via the use of distributed parallel processing, namely artificial neural networks. Four models for describing non-linear MIMO dynamical systems are presented. Based on these models a combined feedforward and recurrent neural networks are structured to emulate the dynamical system. Further, a procedure to emulate multiple systems is suggested. A method for finding a minimal realization of a network is introduced. The minimization greatly reduces the complexity of the network without degrading the operating performance of the network. This work also examines the application of artificial neural networks for adaptive control. The multiple system approach is used to find an adaptive neural network controller for non-linear MIMO time-varying system in a direct model reference control scheme. The controller network is trained using a procedure called back-propagation through the plant, which was extended in this work. The application of neural networks is demonstrated on a longitudinal model of the F/A-18A fighter aircraft both with the undamaged aircraft and with a damage mechanism as a time-varying MIMO dynamical system.

A methodology was developed for manually training autonomous control systems based on artificial neural systems (ANS). In applications where the rule set governing an expert's decisions is difficult to formulate, ANS can be used to extract rules by associating the information an expert receives with the actions taken. Properly constructed networks imitate rules of behavior that permits them to function autonomously when they are trained on the spanning set of possible situations. This training can be provided manually, either under the direct supervision of a system trainer, or indirectly using a background mode where the networks assimilates training data as the expert performs its day-to-day tasks. To demonstrate these methods, an ANS network was trained to drive a vehicle through simulated freeway traffic.