Download Data Driven Predictions of Process Variables [for nuclear PDF

Read or Download Data Driven Predictions of Process Variables [for nuclear powerplants] PDF

Best nuclear books

All About History: The World’s Worst Disasters

All approximately background is the stunningly realised new journal from the makers of ways it really works and All approximately area. that includes attractive illustrations, images and snap shots depicting every little thing from old civilisations to the chilly battle, All approximately heritage is on the market and pleasing to all and makes historical past enjoyable for the complete kin.

Advances in Nuclear Science and Technology, Volume 23

Quantity 23 makes a speciality of perturbation Monte Carlo, non-linear kinetics, and the move of radioactive fluids in rocks.

Nuclear and Particle Physics: The Changing Interface

"Nuclear and Particle Physics" either were very designated topics for many years, and at the moment are constructing increasingly more interfaces. therefore, hitherto common equipment of particle physics are followed by means of nuclear physics. The authors attempt to construct bridges among either fields and provides nuclear physicists a radical advent from the basics of particle physics to present study during this box.

Additional info for Data Driven Predictions of Process Variables [for nuclear powerplants]

Example text

12. The error (%) between the predicted values and the measurements for the level Controller Output as a function of Bypass MOV, Water Level and Outlet flow rate. 13. The GMDH predicted values (“o”) against the test output values (x) for the Tank Water Level as a function of Inlet Flow rate and Outlet MOV positions. 14. The GMDH predicted values (“o”) against the training output values (“x”) for the Tank Water Level as a function of Inlet Flow rate and Outlet MOV positions. 15. The error (%) between the predicted values and the measurements for the Tank Water Level as a function of Inlet Flow rate and Outlet MOV positions.

This case occurs when we have Outlet MOV position fault (single fault) or Water Level sensor fault and Outlet MOV position 38 fault (dual fault). In both these cases, the values of the GMDH predictions have similar behavior. In such cases, additional information or measurements are necessary to resolve the faulty situation. 6 Results of Simulation of Fault Detection and Isolation Studies The following is a summary of the results of a simulation study of the fault detection and isolation simulation performed using the Matlab-Simulink model.

This variable is used later to recover the best of the GMDH models. The loop runs 2n times, where n is the number of basic functions to be tested. At the end of this process, the variable called RESIDUAL contains the residuals of all possible combinations of basic functions (2n). Using the Matlab function called SORT ([y, index] = sort(residual)) the residuals are sorted from smallest to the highest values in the variable Y. The INDEX variable contains the JobCounter case numbers associated with the residuals in Y.

Download PDF sample

Rated 4.42 of 5 – based on 28 votes