ANOOMALY DETECTION SYSTEM AND METHOD FOR ELECTRIC DRIVES

• US 20200132773A1
• Filed: 10/31/2019
• Published: 04/30/2020
• Est. Priority Date: 10/31/2018
• Status: Active Grant

• Current Assignee: ABB Schweiz AG (ABB Limited)
• Original Assignee: ABB Schweiz AG (ABB Limited)
• Inventors: Gugaliya, Jinendra, Kumar-Vij, Rahul, Pani, Abhilash, Sisodia, Arpit, Podila, Nikhil Venkata Saisantosh
• Granted Patent: US 11,467,214 B2
• Time in Patent Office: Days
• CPC Class Codes: G01R 31/343 in operation
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/01 Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
• CPC Class Codes: G07C 3/005 during manufacturing process
• CPC Class Codes: G07C 3/14 Quality control systems

Abstract

One illustrative method comprises retrieving a set of drive information associated with an operation of an electric drive in a time period from a drive control unit and one or more sensor units. The method further includes obtaining reliability information associated with the electric drive from a server. The reliability information is computed in the server, with the set of drive information and a model corresponding to a drive type of the electric drive. The reliability information includes a probability of occurrence of the abnormal condition in a specified future time period. The method further includes, providing at least one condition indication regarding the abnormal condition to one of a human machine interface and a drive controller based on the probability of occurrence of the abnormal condition.

Claims

1. A method for monitoring a condition of an electric drive in a process plant, the electric drive for controlling operation of an electric motor, the method comprising:

obtaining, with the electric drive from at least one of a drive control unit and one or more sensor units, a set of drive information associated with an operation of the electric drive in a time period and operating status information of the electric motor in the time period, wherein the set of drive information comprises values of operating parameters of the electric drive;

obtaining, with the electric drive from a server, reliability information associated with the electric drive, wherein the reliability information comprises a probability of occurrence of an abnormal condition in a future time period, wherein the reliability information is computed in the server using the set of drive information and a model corresponding to the drive type of the electric drive; and

providing, with the electric drive to one of a human machine interface and a drive controller, at least one condition indication regarding the abnormal condition based on the probability of occurrence of the abnormal condition.

2. The method of claim 1, wherein the operating status information of the electric motor is obtained from a status indicator associated with the electric drive, wherein the operating status information comprises one or more time instances associated with one of an energized state, an idle state, and a de-energized state of the electric motor during the time period, and wherein one or more values of the operating parameters at time instances when the electric motor is in the de-energized state are excluded from the values of the operating parameters during computation of the reliability information in the server.

3. The method of claim 2, wherein obtaining the reliability information further comprises obtaining one or more values of one or more operating parameters affecting the abnormal condition, wherein a value of an operating parameter is determined to affect the abnormal condition in the server using the model corresponding to the drive type.

4. The method of claim 3, wherein the reliability information further comprises a condition of the electric drive at a present instant, one or more abnormal operating instances identified during the time period, an expected time for occurrence of the abnormal condition, and at least one recommendation corresponding to an operation of the electric drive.

5. The method of claim 4, wherein the at least one condition indication is provided when the probability of occurrence of the abnormal condition is greater than a first threshold, wherein the probability of occurrence of the abnormal condition is greater than the first threshold when a number of abnormal operating instances identified during the time period is greater than a second threshold.

6. The method of claim 3, wherein providing the at least one condition indication further comprises providing one or more of an alert and a recommendation when the at least one condition indication is provided to the HMI, wherein the recommendation depends on the one or more operating parameters that affected the abnormal condition.

7. The method of claim 3, wherein providing the at least one condition indication further comprises providing at least one recommendation to the drive controller, wherein the drive controller initiates corrective action to prevent occurrence of the abnormal condition based on the at least one recommendation, wherein the at least one recommendation depends on the one or more operating parameters that affected the abnormal condition.

8. The method of claim 1, wherein the model is configured from history data associated with an operation of a plurality of electric drives in a plurality of process plants, wherein the history data of each electric drive of the plurality of electric drives comprises a plurality of drive information obtained over a period of time, wherein the plurality of electric drives correspond to the drive type of the electric drive.

9. Apparatus comprising:

at least one electric drive configured to control operation of an electric motor in a process plant, wherein the at least one electric drive comprises;

one or more sensor units configured to measure values of operating parameters of the at least one electric drive,

a drive control unit configured to determine a set of drive information associated with the at least one electric drive, wherein the set of drive information comprises the values of the operating parameters,

a data acquisition unit to obtain the set of drive information from at least one of the drive control unit and the one or more sensor units. and

a communication unit configured to communicate at least one condition indication to at least one human machine interface, the at least one condition indication being based on a probability of occurrence of an abnormal condition in a future time period; and

a server coupled to the electric drive, the server comprising;

an input interface configured to receive the set of drive information for a time period from the communication unit of the at least one electric drive,

a computing unit configured to compute reliability information associated with the at least one electric drive, based on a model corresponding to a drive type of the at least one electric drive selected by the computing unit from among a plurality of models corresponding to different drive types, wherein the reliability information comprises the probability of occurrence of the abnormal condition in the future time period, a condition of the at least one electric drive at a present instant, one or more abnormal operating instances identified during the time period, and an expected time for occurrence of the abnormal condition, and

an output interface configured to communicate the computed reliability information to the communication unit of the at least one electric drive.

10. The apparatus of claim 9, wherein the computing unit of the server is configured to create each of the plurality of models by applying a machine learning method to history data associated with operation of a plurality of electric drives corresponding to a drive type, wherein the history data comprises a plurality of drive information received by the communication unit of the at least one electric drive over a period of time.

11. The apparatus of claim 10, wherein the computing unit is further configured to exclude one or more values of the operating parameters prior and post one or more time instances associated with one or more drive trips and drive failures of the at least one electric drive during the creation of each model, wherein the one or more time instances of the one or more drive trips and drive failures are part of the history data;

and

12. The apparatus of claim 11, wherein the computing unit is further configured to exclude one or more values of the operating parameters associated with a de-energized state of the at least one electric drive during the creation of each model, wherein time instances of the de-energized state of the at least one electric drive are obtained as part of the history data.