next up previous
Next: Fundamental Statistic Hypothesis Testing Up: Practical Results Analysis under Previous: RTO Loop under Uncertainties

Practical Results Analysis Procedure

RTO predictions can be severely affected by various uncertainties around RTO loop. Results analysis aims at distinguishing different RTO predictions variability so as to implement the meaningful RTO predictions to track the true plant optimal operating point, whereas reject the unnecessary setpoints change for reducing the RTO loop instability. Consider an RTO system in the presence of trust-region constraints, a practical results analysis procedure is presented in this section based on the previous work (Forbes and Marlin, 1996; Miletic and Marlin, 1998), which involve two steps of tests for:

1)
Fundamental Statistic Hypothesis Testing; and

2)
Auxiliary Lagrangian Multipliers Testing.





Guansong Zhang
Wed Mar 10 15:08:26 EST 1999