	US 12,085,929 B2
	Machine learning based direct method of determining status of facility control loop components
Sanjay Kantilal Dave, Bengaluru (IN); Akanksha Jain, Bhopal (IN); Viraj Srivastava, New Delhi (IN); and Vijoy Akavalappil, Thrissur (IN)
Assigned to Honeywell International Inc., Charlotte, NC (US)
Filed by Honeywell International Inc., Morris Plains, NJ (US)
Filed on Jun. 29, 2021, as Appl. No. 17/361,990.
Claims priority of provisional application 63/045,751, filed on Jun. 29, 2020.
Prior Publication US 2021/0405631 A1, Dec. 30, 2021
Int. Cl. G05B 23/02 (2006.01); G06N 3/049 (2023.01); G06N 3/08 (2023.01)

CPC G05B 23/024 (2013.01) [G05B 23/0272 (2013.01); G06N 3/049 (2013.01); G06N 3/08 (2013.01)]

20 Claims

1. A process comprising:

accessing, by a computer processor, a set of time series production data representative of a control process within a facility control loop comprising at least one control valve;

processing, by computer processor, the set of time series production data using a trained machine-learning algorithm, the trained machine-learning algorithm trained using a training data set comprising positive training data representative of a normal operation of one or more control valves within the facility control loop and negative training data representative of an abnormal operation of one or more control valves within the facility control loop, wherein the training data set comprises actual training data augmented based at least in part on an augmented training data set;

identifying, by the computer processor, one or more abnormalities associated with the at least one control valve in the facility control loop based on output of the trained machine-learning algorithm, wherein the one or more abnormalities associated with the at least one control valve in the facility control loop are indicative of at least a control valve nonlinearity-based abnormality; and

transmitting, by the computer processor, a signal to a computer display device indicating the one or more abnormalities.