Automatic optimization of continuous processes
A system, method, and computer-readable medium are disclosed performing an optimization operation. The optimization operation optimizes continuous processes by identifying process states associated with specific ranges for a limited subset of control parameter inputs. In certain embodiments, the optimization operation states comprise clear, stable, and robust process states. Such an optimization operation provides a simpler and cost effective means to optimize continuous processes. Additionally, such an optimization operation is applicable more rapidly to a wider range of real-world operational issues as they occur regularly in continuous process scenarios.
1. A computer-implementable method for performing an optimization operation on continuous processes generated from a plurality of system sensors, comprising:
transforming a continuous stream of data comprising process states associated with specific ranges of a subset of control parameter inputs, wherein transforming comprises:
dividing the control parameter inputs into outputs, controllable inputs, and uncontrollable inputs; and
identifying constraints for the outputs, controllable inputs, and uncontrollable inputs;
performing a clustering operation on select uncontrollable inputs, wherein the clustering operation includes dividing the ranges into a plurality of reliable repeatable process states;
identifying reliably repeatable optimal states from the reliable repeatable process states, the identifying the -reliably repeatable optimal states comprising defining a key performance indicator as a continuous process output by applying a recursive-partitioning operation to identify input parameter settings that are associated with a desirable process state, the recursive-partitioning operation comprising identifying surrogate parameters and alternative recursive-partitioning models to obtain a pool of combinations of possible parameter settings associated with an optimized process performance;
characterizing relationships among the control parameter inputs for the reliably repeatable optimal states and for non-optimal states; and
determining operational states for the control parameter inputs based on the characterized relationships;
wherein the operational states describe controllable parameters and settings to control the system.
2. The method of claim 1 , wherein:
the subset of control parameter inputs comprises optimization operation states, the optimization operation states comprising clear, stable, and robust process states.
3. The method of claim 1 , further comprising:
extracting the data comprising the process states.
4. The method of claim 3 , further comprising:
excluding data with data errors from the data comprising the process states; and
identifying an aggregation interval for the extracting.
5. The method of claim 1 , further comprising dividing the control parameter inputs into independent variables and dependent variables, wherein the independent variables comprise controllable inputs.
6. The method of claim 1 , wherein the reliably repeatable optimal states are identified using scatterplots.
7. The method of claim 6 , wherein the scatterplots identify multiple observed optimal states and models in data identified through clustering and recursive-partitioning modeling.
8. The method of claim 7 , further comprising performing an optimal model verification operation, wherein profiles of all inputs are executed over the optimal states and models.
9. The method of claim 7 , further comprising the step of translation of multiple models into performance curves for control systems.
10. The method of claim 1 , further comprising discarding an optimal state associated with an uncontrollable input from the uncontrolled inputs.
11. The method of claim 1 , further comprising:
determining final optimized parameter settings by relating results of application of the recursive partitioning operation back to historical data.
12. A system for performing an optimization operation on continuous processes generated from a plurality of system sensors, the system comprising:
a processor;
a data bus coupled to the processor; and
a non-transitory, computer-readable storage medium embodying computer program code, the non-transitory, computer-readable storage medium being coupled to the data bus, the computer program code interacting with a plurality of computer operations and comprising instructions executable by the processor and configured for:
transforming a continuous stream of data comprising process states associated with specific ranges of a subset of control parameter inputs, wherein transforming comprises:
dividing the control parameter inputs into outputs, controllable inputs, and uncontrollable inputs; and
identifying constraints for the outputs, controllable inputs, and uncontrollable inputs;
performing a clustering operation on select uncontrollable inputs, wherein the clustering operation includes dividing the ranges into a plurality of reliable repeatable process states;
identifying reliably repeatable optimal states from the reliable repeatable process states, the identifying the -reliably repeatable optimal states comprising defining a key performance indicator as a continuous process output by applying a recursive-partitioning operation to identify input parameter settings that are associated with a desirable process state, the recursive-partitioning operation comprising identifying surrogate parameters and alternative recursive-partitioning models to obtain a pool of combinations of possible parameter settings associated with an optimized process performance;
characterizing relationships among the control parameter inputs for the reliably repeatable optimal states and for non-optimal states; and
determining operational states for the control parameter inputs based on the characterized relationships;
wherein the operational states describe controllable parameters and settings to control the system.
13. The system of claim 12 , wherein:
the subset of control parameter inputs comprises optimization operation states, the optimization operation states comprising clear, stable, and robust process states.
14. The system of claim 12 , wherein the computer program code further comprises instructions executable by the processor for:
extracting the data comprising the process states.
15. The system of claim 12 , wherein the computer program code further comprises instructions executable by the processor for:
excluding data with data errors from the data comprising the process states; and
identifying an aggregation interval for the extracting.
16. The system of claim 12 , wherein the computer program code further comprises instructions executable by the processor for:
determining final optimized parameter settings by relating results of application of the recursive partitioning operation back to historical data.
17. A non-transitory, computer-readable storage medium embodying computer program code, the computer program code comprising computer executable instructions configured for:
transforming a continuous stream of data comprising process states associated with specific ranges of a subset of control parameter inputs, wherein transforming comprises:
dividing the control parameter inputs into outputs, controllable inputs, and uncontrollable inputs; and
identifying constraints for the outputs, controllable inputs, and uncontrollable inputs;
performing a clustering operation on select uncontrollable inputs, wherein the clustering operation includes dividing the ranges into a plurality of reliable repeatable process states;
identifying reliably repeatable optimal states from the reliable repeatable process states, the identifying the -reliably repeatable optimal states comprising defining a key performance indicator as a continuous process output by applying a recursive-partitioning operation to identify input parameter settings that are associated with a desirable process state, the recursive-partitioning operation comprising identifying surrogate parameters and alternative recursive-partitioning models to obtain a pool of combinations of possible parameter settings associated with an optimized process performance;
characterizing relationships among the control parameter inputs for the reliably repeatable optimal states and for non-optimal states; and
determining operational states for the control parameter inputs based on the characterized relationships;
wherein the operational states describe controllable parameters and settings to control the system.
18. The non-transitory, computer-readable storage medium of claim 17 , wherein:
the subset of control parameter inputs comprises optimization operation states, the optimization operation states comprising clear, stable, and robust process states.
19. The non-transitory, computer-readable storage medium of claim 17 , wherein the computer program code further comprises computer executable instructions for: extracting the data comprising the process states.
20. The non-transitory, computer-readable storage medium of claim 17 , wherein the computer program code further comprises computer executable instructions for:
excluding data with data errors from the data comprising the process states; and
identifying an aggregation interval for the extracting.
21. The non-transitory, computer-readable storage medium of claim 17 , wherein the computer program code further comprises computer executable instructions for:
determining final optimized parameter settings by relating results of application of the recursive partitioning operation back to historical data.