System and method for power transmission and distribution asset condition prediction and diagnosis

• US 10,001,518 B2
• Filed: 02/04/2013
• Issued: 06/19/2018
• Est. Priority Date: 02/04/2013
• Status: Active Grant

• Current Assignee: Hitachi Energy Limited (Hitachi, Ltd.)
• Original Assignee: ABB Schweiz AG (ABB Limited)
• Inventors: Cheim, Luiz, Lin, Lan
• Primary Examiner(s): Le, Toan
• Application Number:
  US13/759,026
• Publication Number:
  US 20140222355A1
• Time in Patent Office: 1,961 Days
• CPC Class Codes: G01R 31/327 Testing of circuit interrupters, switches or circuit-breakers
• CPC Class Codes: G01R 31/392 Determining battery ageing or deterioration, e.g. state of health
• CPC Class Codes: G01R 31/62 Testing of transformers
• CPC Class Codes: G05B 23/024 Quantitative history assessment, e.g. mathematical relationships between available data; Functions therefor; Principal component analysis [PCA]; Partial least square [PLS]; Statistical classifiers, e.g. Bayesian networks, linear regression or correlation analysis; Neural networks

Abstract

A computer-implemented system and method for predictive and diagnostic analysis of an electrical transmission, generation and distribution asset health includes a computer with a non-transient computer readable medium able to receive data regarding an asset, its components, component subsystem and parameters related thereto. Instructions stored on the non-transient computer readable medium execute instructions that predictively calculate overall asset health and also calculated the states of subsystems and component parameters, providing a diagnostic of the causes of poor asset health.

Claims

1. A detection and diagnostic system for determining a health of a power transformer having one or more subsystems, comprising:

a plurality of online sensors configured to continuously measure characteristics of power transformer; and

a computer having a non-transient computer readable medium and configured to receive measurements from the plurality of online sensors, component data, and at least one of component parameter data and subsystem data, and configured to store, with the non-transient computer readable medium, information including;

the online sensor measurements,

a Bayesian model including an asset probability distribution, a component probability distribution, a component parameter probability distribution and a subsystem probability distribution, and

a set of instructions that upon execution are structured to;

update one of the asset probability distribution, the component probability distribution, the component parameter probability distribution or the subsystem probability distribution using the online sensor measurements, the component data, the component parameter data or the subsystem data,

update the asset probability distribution using the component probability distribution, the component parameter probability distribution or the subsystem probability distribution,

update the component parameter probability distribution using the asset probability distribution for power transformer or the component probability distribution,

update the subsystem probability distribution using the asset probability distribution, the component probability distribution, or the component parameter probability distribution,

determine the health of power transformer based on updated probability distributions,

predict a fault of the power transformer using the updated probability distributions, and

operate a supervisory control and data acquisition (SCADA) system in response to prediction of the fault to perform supervisory control of the power transformer to a state for maintenance of the power transformer by a technician.

2. The system of claim 1, wherein the computer includes a user input interface and is configured to receive at least one of the component data, component parameter data and subsystem data from at least one of the online sensors and the user input interface.

3. The system of claim 1, wherein the subsystem probability distribution or the component parameter probability distribution includes a probability distribution of at least one of an oil quality, an oil leakage, a load or an infrared parameter.

4. The system of claim 1, wherein the component probability distribution is determined based upon the subsystem probability distribution, the component parameter probability distribution, and a data structure associating the subsystem probability distribution and the component parameter probability distribution with the component probability distribution.

5. The system of claim 1 wherein the instructions written upon the non-transient computer readable medium that upon execution update one of the asset probability distribution, the component probability distribution, the component parameter probability distribution or the subsystem probability distribution perform a Bayesian calculation upon execution.

6. The system of claim 1 wherein the asset probability distribution, the component probability distribution, the component parameter probability distribution and the subsystem probability distribution written on the computer readable medium are normalized.

7. The system of claim 1 further comprising:

an instruction of-the set of instructions written upon the non-transient computer readable medium that upon execution updates the asset probability distribution based upon received asset data.

8. The system of claim 1, wherein the probability distribution of the component includes probability distributions of at least one of an oil preservation system, a load tap changer, a cooling system, an active part, a bushing and a main tank.

9. A detection and diagnostic system for determining a health of an electrical transmission or distribution asset having one or more subsystems, comprising:

a plurality of online sensors configured to measure characteristics of the asset; and

a computer having a non-transient computer readable medium and configured to receive measurements from the plurality of online sensors and at least one of subsystem condition data and component parameter condition data, and configured to store, with the non-transient computer readable medium, information including;

the online sensor measurements,

a Bayesian network including a probability distribution of the asset, a probability distribution of an asset component, and a probability distribution of a subsystem parameter written on the computer readable medium,

a first instruction written upon the non-transient computer readable medium that upon execution updates the probability distribution of the asset and the probability distribution of the at least one asset component based upon the online sensor measurements and the received at least one component parameter condition data and subsystem condition data,

a second instruction written upon the non-transient computer readable medium that upon execution updates the probability distribution of the asset using Bayes'"'"' Theorem and one of the probability distribution of the asset or the probability distribution of least one asset component and updates the probability distribution of the subsystem parameter using Bayes'"'"' Theorem and one of the probability distribution of the asset or the probability distribution of least one asset component,

a third instruction written upon the non-transient computer readable medium that upon execution determines the health of the asset based on updated probability distributions,

a fourth instruction written upon the non-transient computer readable medium that upon execution predicts a fault will occur within the asset using the updated probability distributions, and

a fifth instruction written upon the non-transient computer readable medium that upon execution transmits a fault alert to a control system including a supervisory control and acquisition network (SCADA) in response to predicting the fault, the fault alert effective to cause the control system to isolate the asset from a power source.

10. The system of claim 9, wherein the computer is further configured to store, with the non-transient computer readable medium, information including a sixth instruction that upon execution updates the probability distribution of the asset and the probability distribution of the at least one asset component based upon the received at least one of the subsystem condition data and the component parameter condition data.

11. The system of claim 10, wherein the computer includes a user input interface and is configured to receive the at least one of component data and the subsystem condition data and component parameter condition data from at least one of the online sensors and the user input interface.

12. The system of claim 10, wherein updating the probability distribution of the asset component includes updating a probability distribution of at least one of an oil quality, an oil leakage, a load and an infrared parameter.

13. The system of claim 10, wherein updating the probability distribution of the asset and the probability distribution of the at least one asset component includes updating the probability distribution of the asset and the probability distribution of the at least one asset component based upon oil quality data, oil leakage data, load data or infrared parameter data.

14. The system of claim 9, wherein the instructions that upon execution update the probability distribution of the asset and the probability distribution of the components include instructions to update the probability distribution at least one of an oil preservation system, a load tap changer, a cooling system, an active part, a bushing and a main tank.

15. A method for determining a health of an electrical transmission or distribution asset having one or more subsystems, comprising:

providing a plurality of online sensors configured to continuously measure characteristics of the asset;

providing a computer having a non-transient computer readable medium and configured to receive online sensor measurements, component data and at least one of subsystem data and component parameter data;

receiving the component data and the at least one of the subsystem data and the component parameter data;

executing instructions written upon the non-transient computer readable medium that update a Bayesian network model including probability distribution of the asset, a probability distribution of a subsystem and a probability distribution of a component parameter based upon the online sensor measurements and the received component data;

executing instructions written upon the non-transient computer readable medium that update the probability distribution of the asset using Bayes'"'"' Theorem and the probability distribution of the component parameter or the probability distribution of the subsystem;

executing instructions written upon the non-transient computer readable medium that update the probability distribution of the component parameter using Bayes'"'"' Theorem and the probability distribution of the asset or the probability distribution of the subsystem;

executing instructions written upon the non-transient computer readable medium that determine the health of the asset based on the updated probability distributions,

executing instructions written upon the non-transient computer readable medium that predict a fault will occur within the asset using the updated probability distributions; and

executing instructions written upon the non-transient computer readable medium that transmit a fault alert to a supervisory control and acquisition network (SCADA) in response to predicting the fault, the fault alert effective to cause the SCADA network to isolate the asset from a power source.

16. The method of claim 15 further comprising:

executing instructions written upon the non-transient computer readable medium that update the probability distribution of the asset and the probability distribution of the component based upon the received at least one of subsystem data and component parameter data.

17. The system of claim 15, wherein the probability distribution of the subsystem or the probability of the component parameter includes the probability distribution of at least one of an oil quality, an oil leakage, a load and an infrared parameter.