Smart Materials Research 5 mot 7 met ° @ 0 Wo \ ara Foor 7 Hs Des ° oO n-3 m2 @ ® Ficune 7: Wind load estimation from pressure data: the tributary area of floor N was divided into smaller areas; pressure forces acting on. cach smaller atea, A, were calculated based on pressure data at the nearest presse tap, mt. [OX WS aR a ie = ww a eo as os ° al GE un om 0 os OB OB | 02 01 01 0000S an OL 02 Input variable "Ace30" Input vale Ae 30" « ® A eee ee es St sr as as ° ° NM “03-04-03 02-010 01 02 03 04 08 04-03-02 -O1 0 ON 02 03 04 Input varie Acca” Input aie "e748" @ @ Floune &: Membership fanctions forthe input measured accelerations inthe x-direction (Acc-30, Acc-x48) andthe y-direction (Ace-y- 30,Ace-y-48). loads are x and y directions, respectively. 0) = (F(t) Fy(0]", in which F,(t) and B,(¢) are m x 1 vectors of external forces, acting in x and y directions, respectively. Using the first N ‘modes obtained by FEM with the next transformation MX + CX +KX = F(r), w where X = [x y]" is a 2n x 1 vector and m is the number of nodes, while x and y are vectors of nodal displacements, X= 0Q, @‘Smart Materials Research VE NL"NM” NS NVSZRPWS PS” PM” PL PVL “3 2-10 12 3 4 5 xi! ‘Output variable “force” @ NYT NU NM NS NVSZRPVS” Ps PM PL PL os “I 08 06-04-02 0 02 04 06 OR 1 x10 Output variable force y" o Figure 9: Membership functions forthe output contro force inthe x-direction (Force-x) and the y-direction (Force-y). or 0203 Oa Frequency (He) — Acceleration in -ditctons for 90 deg ‘Aceceaton in y-direction for 90 de, os 0807 Ficune 10: Power spectra of the aceleration response ofthe top coer ofthe building inthe two lateral where @ is 2n x N matrix of eigenvectors and Q is Nx 1 vector of generalized displacements; that is, Pili) 2m) ++ Owl) Pilx2) Polar) +++ Pw(a2) Pil%e) rl%e) e+ Pula) dln) Gol) > dw | din) dala) +++ ond C 3 ° $10) 200m) *++ vl) a n 4x Substituting by (2) into (1) and premultiplying by 7, one obtains MOQ +@"COQ+O"KOQ = OF), (4) By assuming the damping matrix, C, to be proportional damping, (4) results into six uncoupled equations mud tends thugs = Soita)Fl) + Yip) Fl) = GF, ‘mai end + kas = Lextorut + Ze.65.A0= Ge 5 imi + end + kung = SOx Fault) + FOv (Bult) GF, where mis cis kin and GF, are generalized mass, generalized damping, generalized stiffness, and generalized force of the ith mode, respectively. Using the measurements obtained by the pressure transducers, pressure coefficients (matrix Cp) are evaluated at each tap location as a function of time. These values are used with the full-scale model to give the pressure distribution on the surface. The pressure values on the surface of the prototype can be calculated as le P( spacetime) = 3 pU°Cp (spacetime), 6) where P(space,time) is a matrix containing the pressure values on the surface of the full scale model as a function of space (x, y, and 2) and time; p isthe air density which is assumed to be 1.25kg/m? (according to [2]), and U is the prototype mean wind speed. The wind load at any node of the outer surface is the integration of the pressure over the surface area in the vicinity ofthe node as Finodes time) = | P(space,time)da 0‘Smart Materials Research ict (contig. 0.2) 1riretion (contig 0.2) 02 02 = é Foss Fos be i ous 05 a ee) Ss hoo Wind diction angle lg) Wind diction ne) ® ® ous $02 = ; E oa 3 2 jos : i Foust | 5 gos i os a a Ss ia nas OS ‘Wind direction angle (deg) ‘Wind divection angle (deg) oo re *r ptes > RTMD (fuzzy) © 16, 19, 22, 26, 29, 32, 35, 38, 41, 4, and 48 in addition to the displacement of the inertial mass of the damper. A is a (52 x 32) system matrix, B isa 32 location vector, and B isa 32exctation vector. In the system reduction above, the wind loads acting on each of the 15 floors above are computed from the wind loads F acting on each of the 48 floors by Jumping wind forces on adjacent floors atthe locations that correspond to the 15 DOF model The controlled output vector, ye, and the measured ‘output, ymu Of the ROS described by (10) can be expressed Ye= C24 Def + Fon You = Cyt+ Daf + By We 4% a where C., De, Fey Gy Dr and Fy, ate matrices with appro- priate dimensions and » is the measurement noise vector. ‘The model used for controller design was further reduced as, follows 2 =A, +B f+ Ewe, Yo = Cote + Def + Fe Ws (a2) Your = C2e + Dine f+ Fre + Yr where 2, isa 6-dimensional state vector of the reduced order, system, yer is a controlled output vector identical of y. oO Sas wo ‘Wind direction angle (deg) =#- Uncontalled ATM (LQR) > TMD (fuzzy) @ ut 14: RMS accelerations ofthe top commer ofthe tower. defined by (11) Yr isthe measured output vector; i the measurement noise, and Cy, Das Fas Gry Dar ad Fp are appropriate matrices 3. Controllers and Limitations In this study, both TMDs and ATMDs are used for the re- duction of the lateral responses of the building. However, in order to make the design of such control systems more reais ticand applicable, the following restrictions and assumption, are applied. (i) The mass ofthe TMD in the x-direction is 100ton, while the mass of the TMD in the y-direction is 150 ton. Such restrictions are applied to avoid exces- sive weight on the roof (the overall mass on the roof is about 0.625% of the overall building’ mas). (ii) The TMDs are tuned to the first vibrational mode in each corresponding lateral direction. The damping, factor is taken to be 20% of the critical. This amount ‘of damping is selected higher than the optimal value forthe sake of restricting the stroke of the ATMDs. (iil) The maximum stroke of the actuators is restricted to 15m,‘Smart Materials Research direction config 0.1) u ‘ieton (contig. no.) os 2 gu = g i a 2 pay al ow 5 We ms 1m 5 90 “9051905270315 30 Wid deton ange (ep) Wind direction angle) « © scion (confi 00.2) ition config 0.2) 7 ied —>@ Ope vole - iran mccared — = a [pteiem one Zor * ? a4 =o a S 1) i Sos Bo tos 505 daa «3 02 © 8 so as aw as wo 0 ao aia) Sao SH Wind decom ange) Wind direction angle ep) © @ Ficune 15: Maximum accelerations ofthe top corner of the tover. (iv) The maximum control force ofthe actuator in the y- direction is restricted to 100KN, and that in the x- direction is restricted to 25 RN. (¥) The computational delay and the sampling rate of the digital controller are 0.001 5. (vi) Three acceleration measurements are available for cach lateral direction (at floor 30, roof, and mass of the TMD) [Note that the tower requited a TMD with heavier mass and ATMD with higher control force in one lateral direction than the other, which was basically attributed to geometry. A lineat-quadratic regulator (LQR) design with output weight- ing is selected to give the desired control force using the MATLAB function (lgrym). The state-feedback law f = Ge; minimizes the cost function 1A) = [eye 6 RAE, «) where G is the feedback gain matrix, z, is a 6-dimensional Sate vector ofthe reduced order sytem, the measured ‘output vector, the symbol (') denotes transpose, and Q and Rare weighting matrices amet suis were performed with various weighting matrices, corresponding to various regulated output vectors The teats of these parametric Sti indented that an effective controler could be designed by selecting a vector of regulated responses (0 inde the velocities ofeach Noor. For comparison reasons, fuzzy logic controllers are used in this study to command the actuators of the ATMDs (see Nguyen etal. [34)). From a design point of view, fuzzy logic controllers do not require the complexity of a traditional control system. The measured accelerations can be used di- rectly as input to the fuzzy controller. The main advantages of using a fuzzy control algorithm are summarized in Bat- tain etal. [35] and Samali et al. [36]. According to Samali etal [36], uncertainties of input data are treated in a much easier way by fuzzy control theory than by classical control theory. Since fuzzy controllers are based on linguistic syn- thesis, they possess inherent robustness. Fuzzy controllers can be easily implemented in a fuzzy chip with immediate reaction time and autonomous power supply. Furthermore, the design of fuzzy controller does not requite state reduc tion or concerning about observers. Only two acceleration ‘measurements were used (floor 30 and roof). ‘The input variables to the fuzzy controller were selected as accelerations of floors 30 and 48 and the output as the control force. The membership functions for the inputs were defined and selected as seven triangles with overlaps as shown in Figure 8, For the output, they were defined and selected as nine triangles with overlaps as shown in Figure 9. ‘The fuzzy variables used to define the fuzzy space are ZR (zero), PVS (positive very small, PS (positive small), PM (positive medium), PL (positive large), PVL (positive very large), NVS (negative very small), NS (negative smal), NM (negative medium), NL (negative large), and NVL (negative