| CPC G06F 18/29 (2023.01) [G06F 17/16 (2013.01); G06F 18/217 (2023.01); G06N 7/01 (2023.01); G06N 20/00 (2019.01)] | 22 Claims |

|
1. A computer implemented method for optimizing operations of a physical plant, the method comprising:
mapping a process flow diagram, of a plant process of the physical plant to a graph comprising nodes and edges,
wherein the nodes represent processes, each associated with a set of control and environmental variable, and the edges represent flows between processes, and
wherein the graph is a two-level fully connected feed-forward network of the physical plant with no skip layer assumption;
receiving and processing historical data collected from sensors of the physical plant associated with each process, wherein the historical data includes real-time sensor measurements of flow rates, temperatures, and pressures of the physical plant;
automatically learning a behavior for each node of the graph using a machine learning platform, wherein the behavior is based at least on the received historical data of the physical plant process;
modeling one or more regression functions for each node using the machine learning platform to predict an output of each of the processes, wherein the one or more regression functions are modeled based on the learned behavior for each node; and
sequentially adjusting controls of the physical plant by generating and transmitting control signals to actuators, including valves, and pumps, of the physical plant to effect changes in flow rates, pressures, and/or temperatures such that acceptable storage inventories are maintained in one or more predetermined time periods, based on the modeling.
|