search

mostafa-razavi / TranSFF

0 stars 0 forks source link

Mie-SWF development results #16

Open mostafa-razavi opened 4 years ago

mostafa-razavi commented 4 years ago

C2 results:

I-7_P-2_3.7755846142099716-140.86116318038015

I-7_P-2_3 7755846142099716-140 86116318038015

I1-I11_C2_uresonly_phia7

C2 Update: It seems like running PSO optimization is far better than simple gradient based methods such as lm and trf. Here is the new results based on PSO:

3.779-141.32

i-18_p-3_C2_s3.7791621621833933-e141.3274721137754

i-18_p-3 target res

Parity plot of Z_score vs U_score:

image

C2 results for 3.779-141.32:

C2

Compared to M2 MiPPE

C2

The slope of Psat line is wrong, which is due to slight difference between low density pressures which add up.

mostafa-razavi commented 4 years ago

C12 results

i-11_p-5_C12_s3.77761805192-e139.879651795_s4.02826629064-e80.0520465525

i-11_p-5 target res

Parity plot of Z_score vs U_score:

image

mostafa-razavi commented 4 years ago

Running:

C2 and C12 are running based on the above parameters. If they look good I'll test C4 and maybe C8.

mostafa-razavi commented 4 years ago

C12 PSO x-x_x-x result:

i-11_p-1_C12_s3.77009097872-e138.352084508_s4.00956133506-e79.7196917111

i-11_p-1 target res

C12 TRF x-x_x-x results:

I13_3.78655714362-138.774917539_4.00596800842-79.7941460245_C12_s3.78655714362-e138.774917539_s4.00596800842-e79.7941460245

I13_3 78655714362-138 774917539_4 00596800842-79 7941460245 target res

C4 PSO x-x_x-x result:

i-13_p-1_C4_s3.796-e139.29_s3.99-e79.91

i-13_p-1 target res

mostafa-razavi commented 4 years ago

C2 PSO x-x results:

i-11_p-1_C2_s3.78071566416-e141.502803975 i-11_p-1 target res