The Neural Network Virtual Instrument (NNVI) can model both linear and nonlinear processes. NNVI can be trained using continuous measurements or Lab Data. One of the most useful applications is for instruments installed in harsh conditions where the life expectancy of an in-situ instrument may be limited. Another is creating a virtual instrument using Lab Data. If there is a correlation between the target measurement and other more robust measurements then a virtual instrument can be created.
The Predictive Process Neural Network Model Base Controller (PPNNMBC) was designed as a viable substitute for 1st principle models when their development cost for highly complex processes can stretch the budget. Neural Network can model nonlinear processes so the need for transitioning between linear models is not needed. Applied in Multiple Input Multiple Output simulation, the forward looking PPNNMBC has been shown to perform similar to 1st principle models within the Neural Network model training space.
Just edit this element to add your own HTML.