General Electric Company

Johan Michael Reimann

Naresh Sundaram Iyer

Romano Patrick

Sameer Vittal

Nurali Virani

Piero Patrone Bonissone

Stefan Joseph Cafaro

Brian Chandler Barr

Marc Thomas Edgar

Feng Xue

James Ryan Reepmeyer

Atanu Talukdar

Guruprasad Srinivasan

Jason Michael Nichols

Hardev Singh

Kotesh Kummamuri Rao

Gagan Adibhatla

Evin Nathaniel Barber

Venkatesh Kattigari Madyastha

Paul Alex Ardis

Jose Mendoza

Jose Leon Vega

Yan Chen

Gary Cristofoli

Zhaoyuan Yang

Rahim Panjwani

Frederick John Menditto, Iii

Aaron James Schmitz

Suchot Kongsomboonvech

Richard Anthony Zelinski

Iii Frederick John Menditto

A control system for active stability management of a compressor element of a turbine engine is provided. In one example aspect, the control system includes one or more computing devices configured to receive data indicative of an operating characteristic associated with the compressor element. For instance, the data can be received from a high frequency sensor operable to sense pressure at the compressor element. The computing devices are also configured to determine, by a machine-learned model, a stall margin remaining of the compressor element based at least in part on the received data. The machine-learned model is trained to recognize certain characteristics of the received data and associate the characteristics with a stall margin remaining of the compressor element. The computing devices are also configured to cause adjustment of one or more engine systems based at least in part on the determined stall margin remaining.

Application of machine learning to process high-frequency sensor signals of a turbine engine

A system and method including receiving a set of deep neural networks (DNN) including DNNs trained with an embedded trojan and DNNs trained without any embedded trojan, each of the trained DNNs being represented by a mathematical formulation learned by the DNNs and expressing a relationship between an input of the DNNs and an output of the DNNs; extracting at least one characteristic feature from the mathematical formulation of each of the trained DNNs; statistically analyzing the at least one characteristic feature to determine whether there is a difference between the DNNs trained with the embedded trojan and the DNNs trained without any embedded trojan; generating, in response to the determination indicating there is a difference, a detector model to execute the statistical analyzing on deep neural networks; and storing a file including the generated detector model in a memory device.

Fast identification of trustworthy deep neural networks

The example embodiments are directed to a system and method for predicting a flow about an object through the use of a predictive model instead of a machine simulation. Traditional CFD simulations can take hours, even days. The example embodiments provide a predictive model that can predict a CFD flow in seconds which greatly improves design time. In one example, the method may include receiving input data comprising shape parameters of a geometric object and flow parameters associated with the geometric object, predicting, via execution of a predictive model, a computational fluid dynamic (CFD) flow about the geometric object based on the shape parameters and the flow parameters included in the input data, and outputting one or more attributes of the predicted CFD flow about the geometric object via a display device.

Deep learning surrogate for turbulent flow

According to some embodiments, a system, method and non-transitory computer-readable medium are provided comprising a Hypothesis Generation Engine (HGE) to receive one or more property target values for a material; a memory for storing program instructions; an HGE processor, coupled to the memory, and in communication with the HGE, and operative to execute program instructions to: receive the one or more property target values for the material; analyze the one or more property target values as compared to one or more known values in a knowledge base; generate, based on the analysis, an initial set of hypothetical structures, wherein each hypothetical structure includes at least one property target value; execute a likelihood model for each candidate material to generate a likelihood probability for each hypothetical structure, wherein the likelihood probability is a measure of the likelihood that the hypothetical structure will have the target property value; convert each hypothetical structure into a natural language representation; execute an abduction kernel on the natural language representation with the at least one likelihood probability, to output at least one proposed structure that satisfies a likelihood threshold for having the property target value; and receive the output of the executed abduction kernel at a testing module to determine whether the output satisfies the property target values. Numerous other aspects are provided.

Autonomous reasoning and experimentation agent for molecular discovery

A controller includes a processor and memory. The memory stores instructions that, when executed, are configured to cause the processor to receive measurements pertaining to a measured operation parameter of at least a portion of a turbine system. The instructions are also configured to cause the processor to generate a customized model for the at least the portion of the turbine system. Moreover, the instructions are configured to cause the processor to estimate an estimated value using the received measurements. The estimated value pertains to a parameter of the turbine system. Furthermore, the instructions are configured to cause the processor to using the customized model, reduce or remove at least some environmental conditions from a corrected estimated value derived from the estimated value.

Model-based performance estimation

A method, system, and non-transitory computer-readable medium, the method including determining automatically, by a processor, whether behavior for a model representing a plurality of entities and relationships therebetween deviates from a reference behavior for the model; determining, in response to the determination that the model does deviate from the reference behavior, at least one basis for the deviation; automatically forecasting an estimate of a remaining useful life for the model; and modifying the model to compensate for the deviation by at least one of modifying the model to accommodate the deviation and updating the model based on at least one new requirement.

Method and system to automate the maintenance of data-driven analytic models

An automated prognostics system includes a sensor system configured to obtain measurement data by monitoring one or more parameters at one or more locations on each of one or more movable components of an object when the object is subjected to a first mode of operation. The system also includes a computing device having an input interface and a processor. The input interface receives input information such as spatial coordinates information associated with the one or more locations, the measurement data obtained by the sensor system, operational data associated with the first mode of operation, and structural data associated with the object. The processor processes the input information and generates a prognostics report on at least a first component of the one or more movable components, the prognostics report including at least one of a failure prognostic or a likelihood-of-failure prognostic of the first component.

Automated prognostics systems and methods

An automated analytics system can include a sensor system that obtains measurement data by monitoring one or more parameters at each of a number of locations on each of a number of replicated components of an object. A computing device receives the measurement data from the sensor system and uses the measurement data to automatically generate a computerized representation of each of the plurality of replicated components. Thereafter, upon receipt of an input query, the computing device generates a synthesized representation of the object that is specifically directed to a parameter of interest indicated in the query. The synthesized representation may be displayed in a visual format that is interpretable by a human to derive information associated with the parameter of interest.

Automated analytics systems and methods

Growing community of inventors

Clifton Park, NY, United States of America

Johan Michael Reimann