WEIZHONG YAN
Engineering in Clifton Park Center, NY

7395188, Jul 1, 2008

Dec 7, 2006

11/608076

Kai Frank Goebel - Mountain View CA, US
Piero Patrone Bonissone - Schenectady NY, US
Weizhong Yan - Clifton Park NY, US
Neil Holger White Eklund - Schenectady NY, US
Feng Xue - Clifton Park NY, US
Hai Qiu - Clifton Park NY, US

General Electric Company - Niskayuna NY

G06F 11/30, G21C 17/00

702184

A method to predict equipment life is disclosed. The method includes making available a set of input parameters, and defining a model of a health of the equipment as a function of the set of input parameters. The method continues with receiving at least one signal representative of a respective one of an actual sensor output relating to an actual operation attribute margin of the equipment, predicting a remaining useful equipment life based upon a sequence of outputs of the model of the health of the equipment, and generating a signal corresponding to the remaining useful equipment life.

2013013, May 30, 2013

Jan 24, 2013

13/749202

LOCKHEED MARTIN CORPORATION - Bethesda MD, US
Naresh Sundaram Iyer - Ballston Spa NY, US
Weizhong Yan - Clifton Park NY, US

LOCKHEED MARTIN CORPORATION - Bethesda MD

F02C 7/26

60 39091

Starter control valve failure prediction machines, systems, computer readable media, program products, and computer implemented methods to predict and trend starter control valve failures in gas turbine engines using a starter control valve health prognostic and to make predictions of starter control valve failures, are provided. A computer implemented method according to an embodiment of the present invention can include the steps of generating a continuous starter control valve deterioration trend function responsive to a plurality of health indices derived from gas turbine engine startup data downloaded from gas turbine engine sensors for a plurality of startups and analyzing the continuous starter control valve deterioration trend function to identify potential starter control valve failure points where the points on the starter control valve deterioration trend function correlate to a starter control valve health prognostic responsive to historic gas turbine engine startup data downloaded from gas turbine engine sensors.

2004023, Nov 25, 2004

Apr 30, 2003

10/425610

Piero Bonissone - Schenectady NY, US
Rajesh Subbu - Troy NY, US
Weizhong Yan - Clifton Park NY, US
Anindya Chakraborty - Schenectady NY, US

GE Financial Assurance Holdings, Inc.

G06F017/60

705/004000

A method and system for designing a neural network classifier and such a neural network for an automated insurance underwriting system and/or its quality assurance is described. While the design method is demonstrated for quality assurance of automated insurance underwriting, it is broadly applicable to diverse decision-making applications in business, commercial, and manufacturing processes. Specifically, multi-class classification problems are solved by decomposing a multi-class classifier into multiple binary-classifiers, which reduces the complexity of the neural network structure, thus reducing the training time and improving the classification performance. Furthermore, the invention also describes a method to incorporate the domain knowledge into the neural network classifier. Both methods work to improve the performance of the classifier.

7536364, May 19, 2009

Apr 28, 2005

11/116920

Rajesh V. Subbu - Clifton Park NY, US
Piero P. Bonissone - Schenectady NY, US
Neil H. Eklund - Schenectady NY, US
Naresh S. Iyer - Clifton Park NY, US
Rasiklal P. Shah - Latham NY, US
Weizhong Yan - Clifton Park NY, US
Chad E. Knodle - Dayton NV, US
James J. Schmid - Kirkland WA, US

General Electric Company - Schenectady NY

G06N 7/00, G05B 13/04

706 13, 700 28, 700 29, 700 30, 700 32, 700 33, 700 34

A method and system for performing model-based multi-objective asset optimization and decision-making is provided. The method includes building at least two predictive models for an asset. The building includes categorizing operational historical data via at least one of: controllable variables, uncontrollable variables, output objectives, and constraints. The building also includes selecting at least two output objectives or constraints, and identifying at least one controllable or uncontrollable variable suitable for achieving the at least two output objectives or constraints. The method also includes validating each predictive model and performing multi-objective optimization using the predictive models. The multi-objective optimization includes specifying search constraints and applying a multi-objective optimization algorithm. The method further includes generating a Pareto Frontier, and selecting a Pareto optimal input-output vector.

7958062, Jun 7, 2011

May 31, 2007

11/755898

Weizhong Yan - Clifton Park NY, US
Anil Varma - Clifton Park NY, US
Piero Patrone Bonissone - Schenectady NY, US

Lockheed Martin Corporation - Bethesda MD

G06N 5/00

706 12, 706 45

A method and system for creating healthy operating envelope from only data samples obtained during normal operation/behavior of dynamic systems is provided. This method determines healthy operating envelope by clustering a stream of discrete event code sequences from the underlying system under normal operation condition only. The method is unsupervised, that is, requiring no prior knowledge of event code patterns corresponding to different operation conditions. Such created envelope can be used for fault detection and health monitoring of dynamic systems.

2004022, Nov 11, 2004

Apr 23, 2004

10/832003

Richard Messmer - Rexford NY, US
Piero Bonissone - Schenectady NY, US
Kareem Aggour - Schenectady NY, US
Rajesh Subbu - Clifton Park NY, US
Weizhong Yan - Clifton Park NY, US
Naresh Iyer - Clifton Park NY, US

General Electric Company

G06F017/60

705/035000

A system, process and computer program product for underwriting a financial risk instrument application represented by at least one risk attribute is provided. Decision engines examine the at least one risk attribute associated with the financial risk instrument application and assign the application to one of a predetermined set of risk classes. A fusion engine compares the risk classes assigned by each of the decision engines and fuses the assigned risk classes into an aggregated result representative of the risk of the financial risk instrument application. The fusion engine includes a first multi-classifier fusion module that uses an associative function to fuse the assigned risk classes into a first aggregated result and a second multi-classifier fusion that uses a non-associative function to fuse the assigned risk classes into a second aggregated result. A comparison engine selects one of the first aggregated result generated from the first multi-classifier fusion module and the second aggregated result generated from the second multi-classifier fusion module and compares it with a production result generated from the production decision engine. The comparison engine generates an underwriting decision for the financial risk instrument application according to the comparison.

7463465, Dec 9, 2008

Dec 28, 2006

11/646733

Cecil Rivers - Hartford CT, US
Weizhong Yan - Clifton Park NY, US
Yingneng Zhou - Niskayuna NY, US
Xiao Hu - Schenectady NY, US
Sean Dwyer - Milford CT, US
Pradeep Vijayan - Bangalore, IN
Vijaysai Prasad - Bangalore, IN
Karim Younsi - Ballston Lake NY, US

General Electric Company - Schenectady NY

H02H 3/16

361 42, 361 44

A circuit interrupter for interrupting current on a line conductor is provided. The circuit interrupter includes separable contacts, a trip mechanism, a bimetal, a microprocessor, a series arc detection sequence, a low-pass filter circuit, and a high-pass filter circuit. The trip mechanism selectively opens the separable contacts when activated. The series arc detection sequence is resident on the microprocessor and includes a plurality of series fault detection algorithms. The low-pass filter circuit provides a low-pass signal to the series arc detection sequence. The high-pass filter circuit provides a high-pass signal to the series arc detection sequence. The sequence selects a particular algorithm from the plurality of algorithms based on the low-pass signal. The sequence calculates a plurality of statistical features from the high-pass signal and sends an output signal to activate the trip mechanism based on a comparison of the plurality of statistical features to the particular algorithm.

7937334, May 3, 2011

May 31, 2007

11/755924

Piero Patrone Bonissone - Schenectady NY, US
Weizhong Yan - Clifton Park NY, US
Naresh Sundaram Iyer - Clifton Park NY, US
Kai Goebel - Mountain View CA, US
Anil Varma - Clifton Park NY, US

Lockheed Martin Corporation - Bethesda MD

G06F 17/00

706 11, 706 45, 340945

Monitoring dynamic units that operate in complex, dynamic environments, is provided in order to classify and track unit behavior over time. When domain knowledge is available, feature-based models may be used to capture the essential state information of the units. When domain knowledge is not available, raw data is relied upon to perform this task. By analyzing logs of event messages (without having access to their data dictionary), embodiments allow the identification of anomalies (novelties). Specifically, a Normalized Compression Distance (such as one based on Kolmogorov Complexity) may be applied to logs of event messages. By analyzing the similarity and differences of the event message logs, units are identified that did not experience any abnormality (and locate regions of normal operations) and units that departed from such regions.

2004011, Jun 17, 2004

Dec 13, 2002

10/318909

Weizhong Yan - Clifton Park NY, US
Peter Edic - Albany NY, US
Maria Iatrou - Clifton Park NY, US
Kai Goebel - Balston Lake NY, US
Erdogan Cesmeli - Clifton Park NY, US

H05G001/00

378/210000

A technique is provided for minimizing or eliminating motion-related artifacts present in CT images due to the dynamic nature of the imaged tissue or the time-varying concentration of a contrast agent used for imaging. The technique allows for the selection of a region of interest that may encompass a structure of diagnostic interest. For projections acquired by the various rows of detectors in a multi-slice CT imaging system, the portions of the projections attributable to the projection of the region of interest within the detector are averaged for each view angle. The projections containing these averaged values are then combined with the projections which do not encompass the region of interest and the combined projection set may be reconstructed to form a CT image of the dynamic tissue. In addition, a smoothing step may be performed to interpolate projection values around the region of interest to smooth the visual transition to the unaveraged portions of the image.

2012008, Apr 5, 2012

Sep 30, 2010

12/895293

Rajesh Venkat Subbu - Clifton Park NY, US
Lincoln Mamoru Fujita - Roanoke VA, US
Weizhong Yan - Clifton Park NY, US
Noemie Dion Ouellet - Greenville SC, US
Richard J. Mitchell - Greenville SC, US
Piero Patrone Bonissone - Schenectady NY, US
Robert Frank Hoskin - Duluth GA, US

General Electric Company - Schenectady NY

G06F 1/26, G06N 3/02, G06N 3/08

700291, 706 25, 706 21

The present disclosure relates to the use of hybrid predictive models to predict one or more of performance, availability, or degradation of a power plant or a component of the power plant. The hybrid predictive model comprises at least two model components, one based on a physics-based modeling approach and one based on an observational or data-based modeling approach. The hybrid predictive model may self-tune or self-correct as operational performance varies over time.