Neural Networks for Structural Control of a Benchmark Problem, Active Tendon System

Khaldoon Bani-Hani (Civil Engineering, University of Illinois at Urbana-Champaign)
Jamshid Ghaboussi (Civil Engineering, University of Illinois at Urbana-Champaign)

paper (PDF, 754KB); abstract


------------------------------------------------------------

Active Tendon (TEN) Results

  contrl_A
min 1st floor rel. displ.		 contrl_B
min 1st floor abs. accel.		 contrl_C
min both
J1 0.187088 0.154068 0.145416
J2 0.386748 0.330181 0.312098
J3 0.039594 0.036570 0.040954
J4 0.041636 0.034565 0.036030
J5 0.009004 0.008977 0.008722
J6 0.2742660.337341 0.2384170.310296 0.2319600.301091
J7 0.7127190.845894 0.5147670.805162 0.5112290.773103
J8 0.0673620.072086 0.0625520.062222 0.0518850.070818
J9 0.2539790.101358 0.0803880.067414 0.0568700.070769
J10 0.0379250.023808 0.0364140.027949 0.0374040.027260
uRMS 0.701984 0.679113 0.764256
amRMS 2.602144 2.594362 2.520690
xmRMS 0.092650 0.085573 0.095832
uMAX 3.9341942.099568 2.9616011.823422 2.6844392.120308
amMAX 10.9604136.880526 10.5236628.077134 10.8097657.878262
xmMAX 0.4344830.272484 0.4034630.235198 0.3346610.267691
  ElCentroHachinohe ElCentroHachinohe ElCentroHachinohe

Numbers in this face represent violated constraints.
Numbers in this face are the larger of the ElCentro/Hachinohe values.

The Kanai-Tajimi results were performed using wg =14.5 rads/sec, zg =0.3, Tf =750 secs.

------------------------------------------------------------
Page designed by Erik A. Johnson (JohnsonE@usc.edu).