Browse Prior Art Database

METHOD TO DETERMINE THE BEST ECONOMIC TRAJECTORY FOR PLANT OPERATION

IP.com Disclosure Number: IPCOM000250610D
Publication Date: 2017-Aug-08
Document File: 2 page(s) / 8K

Publishing Venue

The IP.com Prior Art Database

This text was extracted from a Microsoft Word document.
This is the abbreviated version, containing approximately 52% of the total text.

METHOD TO DETERMINE THE BEST ECONOMIC TRAJECTORY FOR PLANT OPERATION

This invention determines the setpoints in a plant needed to operate the plant in the most cost-effective (economical) manner, not only for the current production load but for consecutive production loads expected in the near future.  It determines these optimal setpoints by comparing the current setpoints to the most economical ones from historical data that had similar production loads and ambient conditions.

US Patent Application No. 2012/0239167, "Economics-based coordination of advanced process control and real-time optimization":  This approach is a classic implementation of real-time optimization (RTO) combined with advanced process control (APC).  It uses current process and economic data to determine a nonlinear approximation of the economic objective function.  This approximation is then used to provide setpoint targets for the APC system.  However, it does not target the setpoints in the process outside of the APC and it is driven by a steady-state model of the process.

US Patent Application No. 2011/0320388, "System, method and computer program for pattern based intelligent control, monitoring and automation": This invention uses sensor-based process trends and data mining of the same to detect patterns in the process.  These discoveries, however, are not used to drive the towards a better economic point but for developing automated control and monitoring systems for the process.  They are also used for root-cause analysis.

US Patent No. 6,636,842, "System and method for controlling an industrial process utilizing process trajectories":  This invention details a methodology to filter out relevant variables in a process, use a pre-trained neural network model to correlate the relevant variables t...