Temperature based fluid level estimation in an electrical device

Patent metadata

Field	Value
Publication number	US-11437185-B2
Application number	US-202016901115-A
Country	US
Kind code	B2
Filing date	Jun 15, 2020
Priority date	Jun 27, 2019
Publication date	Sep 6, 2022
Grant date	Sep 6, 2022

Abstract

A method and system for predicting performance of a fluid filled electrical device are provided. The system includes a sensing unit operable communicating with a fluid level estimation system. The sensing unit includes one or more sensors physically mountable on and/or around the electrical device, recording temperature data associated with the fluid and the ambient environment. The fluid level estimation system determines temperatures of the fluid and a an ambient temperature, generates feature vectors for one or more of the temperatures based on their correlation with the ambient temperature, and estimates a fluid level inside the electrical device and thereby the performance, based on the feature vectors and a probability density function derived from a distribution constructed using historical temperature gradient data associated with the electrical device.

First claim

The invention claimed is:

1. A system for predicting performance of an electrical device filled with fluid, the system comprising:
a sensing unit configured to record temperature data associated with at least an ambient environment surrounding the electrical device and with fluid filled in the electrical device at pre-determined locations on and around the electrical device; and
a fluid level estimation system in communication with the sensing unit, the fluid level estimation system configured to:
determine from the temperature data, temperatures at the pre-determined locations on and around the electrical device and an ambient temperature for the ambient environment surrounding the electrical device;
generate feature vectors for one or more temperatures using reference temperatures and the ambient temperature; and
predict a level of the fluid inside the electrical device based on the feature vectors and historical temperature gradient data associated with the electrical device.

2. The system of claim 1 , wherein the sensing unit comprises:
at least one first sensor physically mountable on a housing of the electrical device;
at least one second sensor physically mountable on a cooling unit of the electrical device; and
at least one third sensor physically mountable in proximity of the electrical device in the ambient environment surrounding the electrical device.

3. The system of claim 2 , wherein the sensing unit further comprises a sensor physically mountable on a fluid storage component of the electrical device.

4. A fluid level estimation system for predicting a level of the inside an electrical device filled with fluid, the fluid level estimation system comprising:
a non-transitory computer readable storage medium configured to store computer program instructions defined by a plurality of modules of the fluid level estimation system;
at least one processor communicatively coupled to the non-transitory computer readable storage medium, the at least one processor configured to execute the plurality of modules of the fluid level estimation system; and
the plurality of modules of the fluid level estimation system comprising:
a data communication module configured to communicate with a sensing unit comprising a plurality of sensors deployable at pre-determined locations on and around the electrical device filled with fluid, and in proximity of the electrical device, the data communication module further configured to receive temperature data recorded by the plurality of sensors of the sensing unit;
a temperature determination module configured to determine, by processing the temperature data, one or more temperatures associated with the fluid at the pre-determined locations on the electrical device and an ambient temperature of the ambient environment at the pre-determined locations around the electrical device;
a feature vector generation module configured to generate feature vectors for one or more temperatures of the fluid, using a reference temperature of the fluid and the ambient temperature; and
a level prediction module configured to predict a level of the fluid inside the electrical device based on the feature vectors and historical temperature gradient data associated with the electrical device.

5. The fluid level estimation system of claim 4 , wherein the pre-determined locations comprise positions on the electrical device at which the plurality of sensors of the sensing unit are positioned.

6. The fluid level estimation system of claim 4 , wherein the feature vector generation module is further configured to determine a correlation of at least one of the temperatures of the fluid with respect to the ambient temperature, and a correlation of the reference temperature with respect to the ambient temperature, over a time period.

7. The fluid level estimation system of claim 4 , wherein the level prediction module comprises:
a gradient management module configured to generate, using the historical temperature gradient data associated with the electrical device, a distribution of feature vectors for one or more of the temperatures;
a threshold determination module configured to determine a fluid level threshold based on the distribution; and
a parameter comparison module configured to compare the feature vectors with the fluid level threshold.

8. The fluid level estimation system of claim 4 , wherein the level prediction module is configured to initiate one or more notifications based on the level of the fluid predicted, wherein the one or more notifications comprise one or more of a fluid theft alarm, a low fluid level alarm, or a critical fluid level alarm.

9. An electrical device comprising:
a housing;
a core positioned inside the housing, wherein the core is at least partially immersed in a fluid;
at least one cooling unit connected to the housing, the at least one cooling unit configured to cool the fluid; and
a system deployed on and around the electrical device, the system comprising:
a sensing unit configured to record temperature data associated with an ambient environment surrounding the electrical device and with fluid filled in the electrical device at pre-determined locations on and around the electrical device; and
a fluid level estimation system in communication with the sensing unit, the fluid level estimation system configured to:
determine from the temperature data, temperatures at the pre-determined locations on and around the electrical device and an ambient temperature for the ambient environment surrounding the electrical device;
generate feature vectors for one or more temperatures using reference temperatures and the ambient temperature; and
predict a level of the fluid inside the electrical device based on the feature vectors and historical temperature gradient data associated with the electrical device.

10. The electrical device of claim 9 , wherein the electrical device is a fluid cooled electrical device.

11. A method for predicting performance of an electrical device filled with fluid, the method comprising:
receiving from a sensing unit temperature data associated at least with an ambient environment surrounding the electrical device and with fluid filled in the electrical device at pre-determined locations on and around the electrical device;
determining from the temperature data, temperatures at the pre-determined locations on and around the electrical device and an ambient temperature for the ambient environment surrounding the electrical device;
generating feature vectors for one or more temperatures using reference temperatures and the ambient temperature; and
predicting a level of the fluid inside the electrical device based on the feature vectors and historical temperature gradient data associated with the electrical device.

12. The method of claim 11 , wherein generating the feature vectors comprises determining a correlation of at least one of the temperatures with respect to the ambient temperature and a correlation of the reference temperatures with respect to the ambient temperature over a time period.

13. The method of claim 11 , wherein predicting the level of the fluid inside the electrical device comprises:
generating, using the historical temperature gradient data associated with the electrical device, a distribution of feature vectors for one or more of the temperatures;
determining a fluid level threshold based on the distribution; and
comparing the feature vectors with the fluid level threshold.

14. The method of claim 11 , further comprising initiating one or more notifications based on the level of the fluid predicted, wherein the notif

Assignees

• Pradhan Simit
• Siemens Ag

Inventors

• Pradhan Simit

Classifications

• H01F27/402 Primary

  Association of measuring or protective means · CPC title

• G06F17/15 Primary

  Correlation function computation {including computation of convolution operations (arithmetic circuits for sum of products per se, e.g. multiply-accumulators G06F7/5443; digital filters, e.g. FIR, IIR, adaptive filters H03H17/00)} · CPC title

• H02H5/08

  responsive to abnormal fluid pressure, liquid level or liquid displacement, e.g. Buchholz relays · CPC title

• H01F2027/406

  Temperature sensor or protection · CPC title

• H01F2027/404

  Protective devices specially adapted for fluid filled transformers · CPC title