Back to results
US 8,478,568 B2Granted
Prediction system and method
AssigneeGeneral Electric Company
InventorsSreedhar Desabhatla, Shashidhar Nibhanupudi
Life of this patent
Estimating the term…
Abstract
A system for predicting a condition includes an electric machine having slip rings and an exciter providing current and voltage to the electric machine through the slip rings. The system also includes a monitoring device that measures the current and voltage provided by the exciter to the electric machine and that forms a regression model from a current sample and compares values from a next sample to values predicted by the regression model.
Claims19
A system for predicting a condition, the system comprising:
- an electric machine, the electric machine including a rotating shaft and two or more slip rings disposed about the rotating shaft;
- an exciter providing a current and a voltage to the electric machine through the two or more slip rings;
- and
- a monitoring device configured to periodically measure the current and voltage provided by the exciter to the electric machine, the monitoring device further configured to:
- group a first plurality of currents and into a current sample;
- create a regression model from the current sample;
- group a second plurality of currents and voltages into a next sample;
- calculate a series of predicted parameters with the regression model using values from the next sample;
- and
- generate an alarm when the predicted parameters exhibit specific variations from the next sample.
7 more elements
A method of predicting a condition, the method comprising:
- measuring a first plurality of currents and voltages provided from an exciter to a generator;
- storing the first plurality on a computing device as a current sample;
- creating a regression model from the current sample;
- measuring a second plurality of currents and voltages provided from the exciter to the generator;
- storing the second plurality of currents into a next sample;
- calculating on the computing device a series of predicted parameters with the regression model using values from the next sample;
- and
- generating an alarm when the predicted parameters exhibit specific variations from the next sample.
5 more elements
Dependencies are inferred from the claim text.