Optimizing functional operation in manufacturing control

Abstract

A minimal model representation of a manufacturing process describes process behavior for real-time process control and is continuously verified and adjusted. A minimal set of parameters is obtained at any manufacturing level in order to reduce a complex manufacturing process to a manageable form. Contributions of variables to process models are ranked and dynamically partitioned into primary and secondary sets. The primary variables are those which are currently used in the model and are deemed sufficient for full process representation. All other variables are stored in a secondary variable set for possible future transfer to the primary set, if needed, for the process model. As processes change in time, significant secondary variables are automatically transferred by the system to the primary set and insignificant primary variables are transferred to the secondary set. An interactive local and global optimization process is implemented. The local network controls processes locally by driving each of then to its current target point. The global optimization network optimizes the entire set of processes by changing local process target points to values which provide overall product optimization. Operational flow functions are partitioned into five categories: 1) data acquisition and preparation, 2) data evaluation and model testing, 3) improvement of process control and product optimization, 4) detection and treatment of unmeasured but significant process parameters, and 5) process and equipment modeling.

Dependencies are inferred from the claim text.