Slide 1

Objectives

Slide 3

Slide 4

Slide 5

Slide 6

Slide 7

Slide 8

Slide 9

Regularisation

Slide 11

Global Optimisation

Slide 13

Slide 14

Slide 15

Slide 16

Slide 17

Slide 18

Slide 19

Slide 20

Slide 21

Slide 22

Slide 23

Slide 24

Slide 25

Tunneling in parallel

Tunneling in parallel

Why LM-Quasi Newton
 vs T-Gauss-Newton

Slide 29

Another reason to use Truncated Gauss-Newton

Scaling

Scaling

Scaling

The PUNQ Test Case

The PUNQ Test Case

The PUNQ Test Case

Slide 37

PUNQ-10 without noise
initial point = ip       x* = (1, 1, ..., 1)

PUNQ-10 without noise
initial point = ub       x* = (1, 1, ..., 1)

PUNQ-10 with noise
initial point = (1, 1, ..., 1)        x* =  ?

Slide 41

The best minimum ?

Extreme Cases

Scenarios: BHP

Forecast for the Field Properties
FWPT: Field Total Water Production

Forecast Field Gas Total Production

Forecast Field Oil Total production

The error between the real case and the calculated forecast for the 5 minima found.

Slide 49

Conclusions

Conclusions

Conclusions

Slide 53

Future

susanag @ servidor.unam.mx

Truncated Gauss-Newton