Large scale process control by driving factor identification
AssigneeIbex Process Technology, Inc.
InventorsEdward A. Rietman, Jill P. Card
Life of this patent
Estimating the term…
Abstract
Systems and methods of complex process control utilize driving factor identification based on nonlinear regression models and process step optimization. In one embodiment, the invention provides a method for generating a system model for a complex process comprised of nonlinear regression models for two or more select process steps of the process where process steps are selected for inclusion in the system model based on a sensitivity analysis of an initial nonlinear regression model of the process to evaluate driving factors of the process.
Claims20
A method of generating a system model of a process comprising a plurality of sequential process steps, the method comprising the steps of:
- providing an initial process model that has been trained in the relationship between one or more metrics of the process and one or more operational variables of a plurality of process steps;
- providing a process-step model for each of one or more of the plurality of process steps selected based on a sensitivity analysis of the initial process model;
- and
- generating a system model from the one or more process-step models, wherein one or more outputs of the model for one selected process step comprises the input of the model of another selected process step.
A computer-readable medium comprising:
- instructions for providing an initial process model that has been trained in the relationship between one or more metrics of the process and one or more operational variables of a plurality of process steps;
- instructions for providing a process-step model for each of one or more of the plurality of process steps selected based on a sensitivity analysis of the initial process model;
- and
- instructions for generating a system model from the one or more process-step models, wherein one or more outputs of the model for one selected process step comprises the input of the model of another selected process step.
A method for control and optimization of a process, the process comprising a plurality of sequential process steps, the method comprising the steps of:
- providing a system model comprised of one or more process-step models, said process steps being selected based on a sensitivity analysis of an initial process model, wherein one or more outputs of the model for one selected process step comprises the input of the model of another selected process step;
- and
- using an optimizer to determine values for one or more operational variables of the selected process steps within an operational variable constraint set that produce at a substantially lowest cost a predicted process metric for the process as close as possible to a target process metric for said process.
A computer-readable medium comprising:
- instructions for providing a system model comprised of one or more process-step models, said process steps being selected based on a sensitivity analysis of an initial process model, wherein one or more outputs of the model for one selected process step comprises the input of the model of another selected process step;
- and
- instructions for using an optimizer to determine values for one or more operational variables of the selected process steps within an operational variable constraint set that produce at a substantially lowest cost a predicted process metric for the process as close as possible to a target process metric for said process.
A system for control and optimization of a process, the process comprising a plurality of sequential process steps effectuated by one or more process tools, the system comprising:
- a process monitor in electronic communication with one or more process tools;
- a data processing device in electronic communication with the process monitor, wherein in said data processing device is adapted to generate a system model of the process comprised of one or more process-step models, said process steps being selected based on a sensitivity analysis of an initial process model, and where one or more inputs to one or more of the models of the selected process steps is based at least in part on information provided by the process monitor to the data processing device;
- and
- a process-tool controller in one or more of mechanical, electrical, or electromechanical communication with one or more process tools and adapted to adjust an operational variable of the one or more process tools in response to a control signal from the data processing device.
Dependencies are inferred from the claim text.