US 12,066,813 B2
Prediction and operational efficiency for system-wide optimization of an industrial processing system
Dzung Tien Phan, Pleasantville, NY (US); Long Vu, Chappaqua, NY (US); and Dharmashankar Subramanian, White Plains, NY (US)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Mar. 16, 2022, as Appl. No. 17/696,840.
Prior Publication US 2023/0297073 A1, Sep. 21, 2023
Int. Cl. G05B 19/4155 (2006.01); G06N 20/00 (2019.01)
CPC G05B 19/4155 (2013.01) [G06N 20/00 (2019.01); G05B 2219/31449 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
learning, using machine learning, a relationship between an input and a set-point of a plurality of processes and an output of a corresponding process;
deriving a regression function for each process based upon historical data;
training an autoencoder for each process based upon the historical data to form a regularizer;
merging the regression functions and regularizers together into a unified optimization problem;
performing system level optimization using the regression functions and regularizers;
determining a set of optimal set-points of a global optimal solution for operating the processes; and
operating an industrial system based on the set of optimal set-points.