Neural-net based unstable machine identification using individual energy functions. [Transient disturbances in power systems

Energy Technology Data Exchange (ETDEWEB)

Djukanovic, M [Institut Nikola Tesla, Belgrade (Yugoslavia); Sobajic, D J; Pao, Yohhan [Case Western Reserve Univ., Cleveland, OH (United States)

1991-10-01

The identification of the mode of instability plays an essential role in generating principal energy boundary hypersurfaces. We present a new method for unstable machine identification based on the use of supervised learning neural-net technology, and the adaptive pattern recognition concept. It is shown that using individual energy functions as pattern features, appropriately trained neural-nets can retrieve the reliable characterization of the transient process including critical clearing time parameter, mode of instability and energy margins. Generalization capabilities of the neural-net processing allow for these assessments to be made independently of load levels. The results obtained from computer simulations are presented using the New England power system, as an example. (author).

Neural-net based real-time economic dispatch for thermal power plants

Energy Technology Data Exchange (ETDEWEB)

Djukanovic, M.; Milosevic, B. [Inst. Nikola Tesla, Belgrade (Yugoslavia). Dept. of Power Systems; Calovic, M. [Univ. of Belgrade (Yugoslavia). Dept. of Electrical Engineering; Sobajic, D.J. [Electric Power Research Inst., Palo Alto, CA (United States)

1996-12-01

This paper proposes the application of artificial neural networks to real-time optimal generation dispatch of thermal units. The approach can take into account the operational requirements and network losses. The proposed economic dispatch uses an artificial neural network (ANN) for generation of penalty factors, depending on the input generator powers and identified system load change. Then, a few additional iterations are performed within an iterative computation procedure for the solution of coordination equations, by using reference-bus penalty-factors derived from the Newton-Raphson load flow. A coordination technique for environmental and economic dispatch of pure thermal systems, based on the neural-net theory for simplified solution algorithms and improved man-machine interface is introduced. Numerical results on two test examples show that the proposed algorithm can efficiently and accurately develop optimal and feasible generator output trajectories, by applying neural-net forecasts of system load patterns.

Neural-net based coordinated stabilizing control for the exciter and governor loops of low head hydropower plants

Energy Technology Data Exchange (ETDEWEB)

Djukanovic, M.; Novicevic, M.; Dobrijevic, D.; Babic, B. [Electrical Engineering Inst. Nikola Tesla, Belgrade (Yugoslavia); Sobajic, D.J. [Electric Power Research Inst., Palo Alto, CA (United States); Pao, Y.H. [Case Western Reserve Univ., Cleveland, OH (United States)]|[AI WARE, Inc., Cleveland, OH (United States)

1995-12-01

This paper presents a design technique of a new adaptive optimal controller of the low head hydropower plant using artificial neural networks (ANN). The adaptive controller is to operate in real time to improve the generating unit transients through the exciter input, the guide vane position and the runner blade position. The new design procedure is based on self-organization and the predictive estimation capabilities of neural-nets implemented through the cluster-wise segmented associative memory scheme. The developed neural-net based controller (NNC) whose control signals are adjusted using the on-line measurements, can offer better damping effects for generator oscillations over a wide range of operating conditions than conventional controllers. Digital simulations of hydropower plant equipped with low head Kaplan turbine are performed and the comparisons of conventional excitation-governor control, state-space optimal control and neural-net based control are presented. Results obtained on the non-linear mathematical model demonstrate that the effects of the NNC closely agree with those obtained using the state-space multivariable discrete-time optimal controllers.

Neural-net disruption predictor in JT-60U

International Nuclear Information System (INIS)

Yoshino, R.

2003-01-01

The prediction of major disruptions caused by the density limit, the plasma current ramp-down with high internal inductance l i , the low density locked mode and the β-limit has been investigated in JT-60U. The concept of 'stability level', newly proposed in this paper to predict the occurrence of a major disruption, is calculated from nine input parameters every 2 ms by the neural network and the start of a major disruption is predicted when the stability level decreases to a certain level, the 'alarm level'. The neural network is trained in two steps. It is first trained with 12 disruptive and six non-disruptive shots (total of 8011 data points). Second, the target output data for 12 disruptive shots are modified and the network is trained again with additional data points generated by the operator. The 'neural-net disruption predictor' obtained has been tested for 300 disruptive shots (128 945 data points) and 1008 non-disruptive shots (982 800 data points) selected from nine years of operation (1991-1999) of JT-60U. Major disruptions except for those caused by the -limit have been predicted with a prediction success rate of 97-98% at 10 ms prior to the disruption and higher than 90% at 30 ms prior to the disruption while the false alarm rate is 2.1% for non-disruptive shots. This prediction performance has been confirmed for 120 disruptive shots (56 163 data points), caused by the density limit, as well as 1032 non-disruptive shots (1004 611 data points) in the last four years of operation (1999-2002) of JT-60U. A careful selection of the input parameters supplied to the network and the newly developed two-step training of the network have reduced the false alarm rate resulting in a considerable improvement of the prediction success rate. (author)

Evaluating Sparse Linear System Solvers on Scalable Parallel Architectures

National Research Council Canada - National Science Library

Grama, Ananth; Manguoglu, Murat; Koyuturk, Mehmet; Naumov, Maxim; Sameh, Ahmed

2008-01-01

.... The study was motivated primarily by the lack of robustness of Krylov subspace iterative schemes with generic, black-box, pre-conditioners such as approximate (or incomplete) LU-factorizations...

Proto-experiences and subjective experiences: classical and quantum concepts.

Science.gov (United States)

Vimal, Ram Lakhan Pandey

2008-03-01

Deterministic reductive monism and non-reductive substance dualism are two opposite views for consciousness, and both have serious problems. An alternative view is needed. For this, we hypothesize that strings or elementary particles (fermions and bosons) have two aspects: (i) elemental proto-experiences (PEs) as phenomenal aspect, and (ii) mass, charge, and spin as material aspect. Elemental PEs are hypothesized to be the properties of elementary particles and their interactions, which are composed of irreducible fundamental subjective experiences (SEs)/PEs that are in superimposed form in elementary particles and in their interactions. Since SEs/PEs are superimposed, elementary particles are not specific to any SE/PE; they (and all inert matter) are carriers of SEs/PEs, and hence, appear as non-experiential material entities. Furthermore, our hypothesis is that matter and associated elemental PEs co-evolved and co-developed into neural-nets and associated neural-net PEs (neural Darminism), respectively. The signals related to neural PEs interact in a neural-net and neural-net PEs emerges from random process of self-organization. The neural-net PEs are a set of SEs embedded in the neural-net by a non-computational or non-algorithmic process. The non-specificity of elementary particles is transformed into the specificity of neural-nets by neural Darwinism. The specificity of SEs emerges when feedforward and feedback signal interacts in the neuropil and are dependent on wakefulness (i.e., activation) attention, re-entry between neural populations, working memory, stimulus at above threshold, and neural net PE signals. This PE-SE framework integrates reductive and non-reductive views, complements the existing models, bridges the explanatory gaps, and minimizes the problem of causation.

Constructing Frozen Jacobian Iterative Methods for Solving Systems of Nonlinear Equations, Associated with ODEs and PDEs Using the Homotopy Method

Directory of Open Access Journals (Sweden)

Uswah Qasim

2016-03-01

Full Text Available A homotopy method is presented for the construction of frozen Jacobian iterative methods. The frozen Jacobian iterative methods are attractive because the inversion of the Jacobian is performed in terms of LUfactorization only once, for a single instance of the iterative method. We embedded parameters in the iterative methods with the help of the homotopy method: the values of the parameters are determined in such a way that a better convergence rate is achieved. The proposed homotopy technique is general and has the ability to construct different families of iterative methods, for solving weakly nonlinear systems of equations. Further iterative methods are also proposed for solving general systems of nonlinear equations.

Neural Network for Image-to-Image Control of Optical Tweezers

Science.gov (United States)

Decker, Arthur J.; Anderson, Robert C.; Weiland, Kenneth E.; Wrbanek, Susan Y.

2004-01-01

A method is discussed for using neural networks to control optical tweezers. Neural-net outputs are combined with scaling and tiling to generate 480 by 480-pixel control patterns for a spatial light modulator (SLM). The SLM can be combined in various ways with a microscope to create movable tweezers traps with controllable profiles. The neural nets are intended to respond to scattered light from carbon and silicon carbide nanotube sensors. The nanotube sensors are to be held by the traps for manipulation and calibration. Scaling and tiling allow the 100 by 100-pixel maximum resolution of the neural-net software to be applied in stages to exploit the full 480 by 480-pixel resolution of the SLM. One of these stages is intended to create sensitive null detectors for detecting variations in the scattered light from the nanotube sensors.

The return of the whole organism

Indian Academy of Sciences (India)

Unknown

literature on behavioural development for a long time. The importance becomes .... fied metabolism, helping the child to cope with a short- age of food. ..... Bateson P and Horn G 1994 Imprinting and recognition memory – a neural-net model; ...

Unsymmetric ordering using a constrained Markowitz scheme

Energy Technology Data Exchange (ETDEWEB)

Amestoy, Patrick R.; Xiaoye S.; Pralet, Stephane

2005-01-18

We present a family of ordering algorithms that can be used as a preprocessing step prior to performing sparse LU factorization. The ordering algorithms simultaneously achieve the objectives of selecting numerically good pivots and preserving the sparsity. We describe the algorithmic properties and challenges in their implementation. By mixing the two objectives we show that we can reduce the amount of fill-in in the factors and reduce the number of numerical problems during factorization. On a set of large unsymmetric real problems, we obtained the median reductions of 12% in the factorization time, of 13% in the size of the LU factors, of 20% in the number of operations performed during the factorization phase, and of 11% in the memory needed by the multifrontal solver MA41-UNS. A byproduct of this ordering strategy is an incomplete LU-factored matrix that can be used as a preconditioner in an iterative solver.

Fast neural-net based fake track rejection in the LHCb reconstruction

CERN Document Server

De Cian, Michel; Seyfert, Paul; Stahl, Sascha

2017-01-01

A neural-network based algorithm to identify fake tracks in the LHCb pattern recognition is presented. This algorithm, called ghost probability, retains more than 99 % of well reconstructed tracks while reducing the number of fake tracks by 60 %. It is fast enough to fit into the CPU time budget of the software trigger farm and thus reduces the combinatorics of the decay reconstructions, as well as the number of tracks that need to be processed by the particle identification algorithms. As a result, it strongly contributes to the achievement of having the same reconstruction online and offline in the LHCb experiment in Run II of the LHC.

Neural-net predictor for beta limit disruptions in JT-60U

International Nuclear Information System (INIS)

Yoshino, R.

2005-01-01

Prediction of major disruptions occurring at the β -limit for tokamak plasmas with a normal magnetic shear in JT-60U was conducted using neural networks. Since no clear precursors are generally observed a few tens of milliseconds before the β -limit disruption, a sub-neural network is trained to output the value of the β N limit every 2 ms. The target β N limit is artificially set by the operator in the first step to train a network with non-disruptive shots as well as disruptive shots, and then in the second step the target limit is modified using the β N limit output from the trained network. The adjusted target greatly improves the consistency between the input data and the output. This training, the 'self-teaching method', has greatly reduced the false alarm rate triggered for non-disruptive shots. To improve the prediction performance further, the difference between the output β N limit and the measured β N , and 11 parameters, are inputted to the main neural network to calculate the 'stability level'. The occurrence of a major disruption is predicted when the stability level decreases to the 'alarm level'. Major disruptions at the β -limit have been predicted by the main network with a prediction success rate of 80% at 10 ms prior to the disruption while the false alarm rate is lower than 4% for non-disruptive shots. This 80% value is much higher than that obtained for a network trained with a fixed target β N limit set to be the maximum β N observed at the start of a major disruption, lower than 10%. A prediction success rate of 90% with a false alarm rate of 12% at 10 ms prior to the disruption has also been obtained. This 12% value is about half of that obtained for a network trained with a fixed target β N limit

HAWC Analysis of the Crab Nebula Using Neural-Net Energy Reconstruction

Science.gov (United States)

Marinelli, Samuel; HAWC Collaboration

2017-01-01

The HAWC (High-Altitude Water-Cherenkov) experiment is a TeV γ-ray observatory located 4100 m above sea level on the Sierra Negra mountain in Puebla, Mexico. The detector consists of 300 water-filled tanks, each instrumented with 4 photomuliplier tubes that utilize the water-Cherenkov technique to detect atmospheric air showers produced by cosmic γ rays. Construction of HAWC was completed in March, 2015. The experiment's wide field of view (2 sr) and high duty cycle (> 95 %) make it a powerful survey instrument sensitive to pulsar wind nebulae, supernova remnants, active galactic nuclei, and other γ-ray sources. The mechanisms of particle acceleration at these sources can be studied by analyzing their energy spectra. To this end, we have developed an event-by-event energy-reconstruction algorithm employing an artificial neural network to estimate energies of primary γ rays. The Crab Nebula, the brightest source of TeV photons, makes an excellent calibration source for this technique. We will present preliminary results from an analysis of the Crab energy spectrum using this new energy-reconstruction method. This work was supported by the National Science Foundation.

Pattern recognition neural-net by spatial mapping of biology visual field

Science.gov (United States)

Lin, Xin; Mori, Masahiko

2000-05-01

The method of spatial mapping in biology vision field is applied to artificial neural networks for pattern recognition. By the coordinate transform that is called the complex-logarithm mapping and Fourier transform, the input images are transformed into scale- rotation- and shift- invariant patterns, and then fed into a multilayer neural network for learning and recognition. The results of computer simulation and an optical experimental system are described.

Signal and data processing of small targets 1990; Proceedings of the Meeting, Orlando, FL, Apr. 16-18, 1990

Science.gov (United States)

Drummond, Oliver E.

Various papers on signal and data processing of small targets are presented. Individual topics addressed include: clutter rejection using multispectral processing, new sensor for automatic guidance, Doppler domain localized generalized-likelihood-ratio detector, linear filter for resolution of point sources, analysis of order-statistic filters for robust detection, distribution functions for additive Gaussian and gamma noise, temperature discrimination of closely space objects, knowledge-based tracking algorithm, detecting and tracking low-observable targets using IR, weak target detection using the entropy concept, measurement-based neural-net multitarget tracker, object-track closed-form solution in angle space, passive-sensor data association for tracking. Also discussed are: application of MHT to dim moving targets, automatic static covariance analysis with mathematica, neural network implementation of plot/track association, application of Bayesian networks to multitarget tracking, tracking clusters and extended objects with multiple sensors, target tracking by human operator, finite impulse response estimator, multitarget tracking using an extended Kalman filter, maneuvering target tracking using a two-state model algorithm, mixture reduction algorithms for target tracking in clutter, scene interpretation approach to high-level target tracking.

Results from CrIS-ATMS Obtained Using the AIRS Science Team Retrieval Methodology

Science.gov (United States)

Susskind, Joel; Kouvaris, Louis C.; Iredell, Lena

2013-01-01

AIRS was launched on EOS Aqua in May 2002, together with AMSU-A and HSB (which subsequently failed early in the mission), to form a next generation polar orbiting infrared and microwave atmospheric sounding system. AIRS/AMSU had two primary objectives. The first objective was to provide real-time data products available for use by the operational Numerical Weather Prediction Centers in a data assimilation mode to improve the skill of their subsequent forecasts. The second objective was to provide accurate unbiased sounding products with good spatial coverage that are used to generate stable multi-year climate data sets to study the earth's interannual variability, climate processes, and possibly long-term trends. AIRS/AMSU data for all time periods are now being processed using the state of the art AIRS Science Team Version-6 retrieval methodology. The Suomi-NPP mission was launched in October 2011 as part of a sequence of Low Earth Orbiting satellite missions under the "Joint Polar Satellite System" (JPSS). NPP carries CrIS and ATMS, which are advanced infra-red and microwave atmospheric sounders that were designed as follow-ons to the AIRS and AMSU instruments. The main objective of this work is to assess whether CrIS/ATMS will be an adequate replacement for AIRS/AMSU from the perspective of the generation of accurate and consistent long term climate data records, or if improved instruments should be developed for future flight. It is critical for CrIS/ATMS to be processed using an algorithm similar to, or at least comparable to, AIRS Version-6 before such an assessment can be made. We have been conducting research to optimize products derived from CrIS/ATMS observations using a scientific approach analogous to the AIRS Version-6 retrieval algorithm. Our latest research uses Version-5.70 of the CrIS/ATMS retrieval algorithm, which is otherwise analogous to AIRS Version-6, but does not yet contain the benefit of use of a Neural-Net first guess start-up system

Neural-net based calculation of voltage dips at maximum angular swing in direct transient stability analysis [of power systems

Energy Technology Data Exchange (ETDEWEB)

Djukanovic, M [Inst. ' Nikola Tesla' , Belgrade (Yugoslavia); Sobajic, D J; Pao, Yohhan [Case Western Reserve Univ., Cleveland, OH (United States). Dept. of Electrical Engineering and Applied Physics Case Western Reserve Univ., Cleveland, OH (United States). Dept. of Computer Engineering and Science AI WARE inc., Cleveland, OH (United States)

1992-10-01

In heavily stressed power systems, post-fault transient voltage dips can lead to undesired tripping of industrial drives and large induction motors. The lowest transient voltage dips occur when fault clearing times are less than critical ones. In this paper, we propose a new iterative analytical methodology to obtain more accurate estimates of voltage dips at maximum angular swing in direct transient stability analysis. We also propose and demonstrate the possibility of storing the results of these computations in the associative memory (AM) system, which exhibits remarkable generalization capabilities. Feature-based models stored in the AM can be utilized for fast and accurate prediction of the location, duration and the amount of the worst voltage dips, thereby avoiding the need and cost for lengthy time-domain simulations. Numerical results obtained using the example of the New England power system are presented to illustrate our approach. (Author)

Surge of a Complex Glacier System - The Current Surge of the Bering-Bagley Glacier System, Alaska

Science.gov (United States)

Herzfeld, U. C.; McDonald, B.; Trantow, T.; Hale, G.; Stachura, M.; Weltman, A.; Sears, T.

2013-12-01

Understanding fast glacier flow and glacial accelerations is important for understanding changes in the cryosphere and ultimately in sea level. Surge-type glaciers are one of four types of fast-flowing glaciers --- the other three being continuously fast-flowing glaciers, fjord glaciers and ice streams --- and the one that has seen the least amount of research. The Bering-Bagley Glacier System, Alaska, the largest glacier system in North America, surged in 2011 and 2012. Velocities decreased towards the end of 2011, while the surge kinematics continued to expand. A new surge phase started in summer and fall 2012. In this paper, we report results from airborne observations collected in September 2011, June/July and September/October 2012 and in 2013. Airborne observations include simultaneously collected laser altimeter data, videographic data, GPS data and photographic data and are complemented by satellite data analysis. Methods range from classic interpretation of imagery to analysis and classification of laser altimeter data and connectionist (neural-net) geostatistical classification of concurrent airborne imagery. Results focus on the characteristics of surge progression in a large and complex glacier system (as opposed to a small glacier with relatively simple geometry). We evaluate changes in surface elevations including mass transfer and sudden drawdowns, crevasse types, accelerations and changes in the supra-glacial and englacial hydrologic system. Supraglacial water in Bering Glacier during Surge, July 2012 Airborne laser altimeter profile across major rift in central Bering Glacier, Sept 2011

The EB factory project. I. A fast, neural-net-based, general purpose light curve classifier optimized for eclipsing binaries

International Nuclear Information System (INIS)

Paegert, Martin; Stassun, Keivan G.; Burger, Dan M.

2014-01-01

We describe a new neural-net-based light curve classifier and provide it with documentation as a ready-to-use tool for the community. While optimized for identification and classification of eclipsing binary stars, the classifier is general purpose, and has been developed for speed in the context of upcoming massive surveys such as the Large Synoptic Survey Telescope. A challenge for classifiers in the context of neural-net training and massive data sets is to minimize the number of parameters required to describe each light curve. We show that a simple and fast geometric representation that encodes the overall light curve shape, together with a chi-square parameter to capture higher-order morphology information results in efficient yet robust light curve classification, especially for eclipsing binaries. Testing the classifier on the ASAS light curve database, we achieve a retrieval rate of 98% and a false-positive rate of 2% for eclipsing binaries. We achieve similarly high retrieval rates for most other periodic variable-star classes, including RR Lyrae, Mira, and delta Scuti. However, the classifier currently has difficulty discriminating between different sub-classes of eclipsing binaries, and suffers a relatively low (∼60%) retrieval rate for multi-mode delta Cepheid stars. We find that it is imperative to train the classifier's neural network with exemplars that include the full range of light curve quality to which the classifier will be expected to perform; the classifier performs well on noisy light curves only when trained with noisy exemplars. The classifier source code, ancillary programs, a trained neural net, and a guide for use, are provided.

Advanced combustion technologies for gas turbine power plants

Energy Technology Data Exchange (ETDEWEB)

Vandsburger, U.; Desu, S.B. [Virginia Tech, Blacksburg, VA (United States); Roe, L.A.

1995-10-01

During the second half of fiscal year 1995 progress was made in all three funded subject areas of the project as well as in a new area. Work in the area of mixing and combustion management through flow actuation was transferred into an enclosed facility. Jet mixing in a ducted co-flow was examined. The same jets were also subjected to a strong acoustic field established in the duct. Excitation of the jet with static spatial modes was shown to be effective even in the presence of co-flow and the acoustic field. Only when a wall is placed at the jet exit plane did the acoustic field dominate the jet dispersion (as expected due to reflective boundary conditions and the jet shear layer receptivity). This case is, however, not the most relevant to gas turbine combustors since it precludes co-flow. In the area of combustor testing, the design, fabrication, and assembly of a modular combustor test rig for project has been completed at the University of Arkansas. In the area of high temperature piezoceramic actuator materials development, Sr{sub 2}(Nb{sub x}Ta{sub 1-x}){sub 2}O{sub 7} powders have been synthesized, and bulk samples and thick films sintered. These materials have a curie temperature of about 1400{degrees}C compared with 300{degrees}C for the commercially available PZT. While at room temperature the new materials show a piezoelectric constant (d{sub 33}) which is a factor of 100 lower than PZT, at high temperatures they can exhibit significant action. A new area of non-linear, neural-net based, controllers for mixing and combustion control has been added during the second contract year. This work is not funded by the contract. Significant progress was made in this area. Neural nets with up to 15 neurons in the hidden layer were trained with experimental data and also with data generated using linear stability theory. System ID was performed successfully. The network was then used to predict the behavior of jets excited at other modes not used for the training.