research article novel distributed pzt active vibration ...ectiveness of the proposed active...
TRANSCRIPT
-
Research ArticleNovel Distributed PZT Active Vibration Control Based onCharacteristic Model for the Space Frame Structure
Hua Zhong,1 Yong Wang,2 Hanzheng Ran,1 Qing Wang,1 and Changxing Shao2
1 Institute of Electronic Engineering, China Academy of Engineering Physics, Mianyang 621999, China2Department of Automation, University of Science and Technology of China, Hefei 230027, China
Correspondence should be addressed to Hua Zhong; [email protected]
Received 6 July 2015; Revised 19 September 2015; Accepted 21 September 2015
Academic Editor: Mario Terzo
Copyright Β© 2016 Hua Zhong et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
A novel distributed PZT control strategy based on characteristic model is presented for space frame structure in this paper. It is achallenge to obtain the exact mechanical model for space structure, since it is a coupling MIMO plant with unknown parametersand disturbances.Thus the characteristic modeling theory is adopted to establish the needed model, which can accurately describethe dynamic characteristics of the space frame structure in real time. On basis of this model, a keep tracking controller is designedto suppress the vibration actively. It is shown that the proposed model-free method is very robust and easy to implement. To solvethe complex and difficulty problem on PZT location optimization, an efficient method with modal strain energy and maximumvibration amplitude is proposed. Finally, a simulation study is conducted to investigate the effectiveness of the proposed activevibration control scheme.
1. Introduction
Space structures are usually applied in the aerospace andcivil engineering such as the space boom structures [1].Generally, since space structures are light and flexible, theyare sensitive to vibration which will affect the performanceof space equipment [2]. Therefore, it is necessary to suppressvibration for space structure.
Active vibration control method is a good choice forthe vibration control of space structure. To achieve goodvibration control performance, there are three things to bedone, that is, system modelling, sensors/actuators locationconfiguration, and control strategy design.
Space structureβs model can be divided into two cat-egories: mechanical model and identification model. Themechanical model is usually established with sophisticatedfinite element method [3, 4]. Additionally, it is generally usedin system dynamics analysis and sensors/actuators locationoptimization. Nonetheless, it can not be used to designcontroller directly due to its complexity and inaccuracy. Dueto the tremendous efforts devoted by researchers, valuableresults on identificationmodeling for the space structure have
been obtained. Hwang presented an analytical procedurebased on the Kalman filtering approach to estimate modalloads applied on a structure [5]. Hazra et al. proposed amethod which integrates the blind identification with time-frequency decomposition of signals for flexible structures [6].Schoen et al. presented an identification algorithmwhich uti-lizes modal contribution coefficients to monitor the data col-lection for large flexible space structures [7]. Although someaforementionedmethods have been extensively used in spacestructure identification, there still exist several disadvantages.On the one hand, the system is inherently complicated, muchless the disturbances. On the other hand, the existing resultsonly depend on complex off-line identification methods,which is difficult to grasp the system dynamics in real time.Consequently, it is a challenge to establish themodel for spacestructure with on-line identification method.
For active vibration control of space structure, determina-tion of actuator locations is an important work. If locationsof actuators are not proper, active vibration control maynot suppress the vibration, even it can increase structurevibration. Therefore, optimization of actuatorsβ locations hasattracted many scholars to study. Chen et al. presented
Hindawi Publishing CorporationShock and VibrationVolume 2016, Article ID 5928270, 9 pageshttp://dx.doi.org/10.1155/2016/5928270
-
2 Shock and Vibration
a location optimization approach based on the 1st singularvalue perturbations of observability and controllability [8].Xu and Jiang demonstrated an approach with the control-lability and observability index of the system [4]. Zhao etal. proposed a continuous variable optimization method tosolve the optimal placement of piezoelectric active bars [2].Li and Huang presented a layered optimization strategy toaddress location optimization problem [9].Though the abovemethods can realize location optimization actuators, theoptimization process is complex. Thus, it is needed to find asimple and effective optimization method.
An active vibration controller with simple structure andstrong robustness is necessary. Many scholars have devotedthemselves to solving this problem. Luo et al. designed simplePD controller for vibration attenuation in Hoop truss struc-ture [10]. Abreu et al. used a standard π»
βrobust controller
to suppress structural vibrations [11]. Yang et al. presentedan adaptive fuzzy sliding mode controller for vibration ofa flexible rectangular plate [12]. Lin and Zheng proposed aparallel neurofuzzy control with genetic algorithm tuning forsmart piezoelectric rotating truss structure [13]. de Abreu etal. designed a self-organizing fuzzy controller for vibrationcontrol of smart truss structure [1]. Mahmoodi developed amodified positive position feedback controller for vibrationcontrol [14]. Wilhelm and Rajamani applied LQR controllerto realize multimodal vibration suppression [15]. Amongabove control strategies, simple controllers (such as PID)are with weak robustness. The complex controllers (π»
β
controller, fuzzy controller) have strong robustness, but theyhave difficulty in parameter tuning.
Motivated by the above discussions, a novel distributedPZT control based on characteristic model is presented.Characteristic model for space structure is established toaccurately describe systemβs dynamics in time whose formis simple and characteristic parameters of it can contain theinformation of time delay or high order [16]. Additionally, asimple controller based on characteristic model called keeptracking controller with strong robustness is applied to realizevibration control. Furthermore, a novel actuators locationoptimization method with modal strain energy (MSE) andmaximum vibration amplitude is proposed.
The remainder of the paper is organized as follows.Section 2 establishes the characteristic model for space struc-ture. Section 3 presents the identification of characteristicparameters and control strategy. PZT location optimizationis presented in Section 4. Numerical simulation is shown inSection 5. Conclusion is shown in Section 6.
2. Characteristic Modeling for Space Structure
Thespace frame structure is shown in Figure 1which contains26 nodes and 63 beam elements. Nodes 2, 3, 7, and 8are constrained which are fixed with the floor. Structureexcitation direction is π§ of node 10. Every node has 6 degreesof freedom (DOF).The totalDOF is π.The space structure has63 beam elements. Table 1 shows the connection informationamong nodes. Space structure consists of aluminum whosecross section is rectangular. Table 2 shows the top 10 naturalfrequencies.
X
Y
Z
1 2 3 4
5
678
9
10
11
12
13
14
15
1617
18
19
20
21
22
23
2425
26
Figure 1: The schematic diagram of frame structure.
According to the finite element formulation, the equationof motion for any structure is given by
MXΜ + CXΜ + KX = F, (1)
whereM β RπΓπ is mass matrix, C β RπΓπ is damping matrix,K β RπΓπ is stiffness matrix, F β RπΓ1 is load force vector,and X β RπΓ1, XΜ β RπΓ1, and XΜ β RπΓ1 are structuraldisplacement, velocity, and acceleration vectors, respectively.
With the knowledge of space frame, parameters M, C,and K are slow time-varying. Then, during short samplingtime Ξπ‘ (ms), it can be considered thatM, C, and K are time-invariant. Then the discrete form of motion equation (1) canbe described as
MX (π + 1) = (2M β CΞπ‘ β KΞπ‘2)X (π)
+ (βM + CΞπ‘)X (π β 1) + Ξπ‘2F (π) .(2)
One has the upper equation (2) as
X (π + 1) = G1(π)X (π) + G
2(π)X (π β 1)
+H1(π)F (π) ,
(3)
where G1(π) = 2I
πβ Mβ1CΞπ‘ β Mβ1KΞπ‘2, G
2(π) =
βIπ+ Mβ1CΞπ‘, and H
1(π) = Ξπ‘
2Mβ1. The expression (3)is called characteristic model; G
1(π), G
2(π), and H
1(π) are
characteristic parameters.When Ξπ‘ is very small, the following equation holds:
G1(β) = 2I
π,
G2(β) = βI
π,
H1(β) = 0
πΓπ.
(4)
Further one has the sum of characteristic parameters as
G1(β) + G
2(β) +H
1(β) = I
π. (5)
-
Shock and Vibration 3
Table 1: Beam element connection information.
Beam number Node number1 1-22 2-33 3-44 4-55 5-66 6-77 7-88 8-99 9-110 9-1011 1-1012 2-1013 10-814 10-715 10-316 3-717 3-518 5-719 1-2620 25-1121 11-222 24-1223 12-324 23-1325 13-1426 14-427 15-628 16-1529 22-1630 17-731 21-1732 18-833 20-1834 19-935 19-2636 26-2537 26-1138 1-1139 25-2440 25-1241 11-1242 11-343 24-2344 24-1345 12-1346 12-1447 3-1448 23-2249 13-16
Table 1: Continued.
Beam number Node number50 14-1551 21-2252 21-1653 17-1654 17-1555 7-1556 20-2157 20-1758 18-1759 18-760 19-2061 19-1862 9-1863 24-21
Assuming the output X(π) loops are decoupled, thusG1(π) and G
2(π) are diagonal matrixes, which can be formu-
lated as
G1(π) =
[[[[[[
[
π11
(π) 0 β β β 0
0 π12
(π) β β β 0
.
.
.... d
.
.
.
0 0 β β β π1π
(π)
]]]]]]
]
,
G2(π) =
[[[[[[[[
[
π21
(π) 0 β β β 0
0 π22
(π) β β β ...
.
.
.... d 0
0 0 β β β π2π
(π)
]]]]]]]]
]
.
(6)
The characteristic parameterH1(π) can be written as
H1(π) =
[[[[[[
[
β11
(π) β12
(π) β β β β1π
(π)
β21
(π) β22
(π) β β β β2π
(π)
.
.
.... d
.
.
.
βπ1
(π) βπ2
(π) β β β βππ
(π)
]]]]]]
]
. (7)
According to the expression in (3) and (6)-(7), the πthcharacteristic model is
π₯π(π + 1) = π
1π(π) π₯π(π) + π
2π(π) π₯π(π β 1)
+
π
β
π=1
βππ(π) πΉπ(π) .
(8)
-
4 Shock and Vibration
Table 2: Top 10 natural frequencies.
Order 1 2 3 4 5 6 7 8 9 10Frequency (Hz) 10.03 15.51 18.02 22.82 25.85 27.72 29.35 31.45 42.22 42.61
PZT1
PZT2
PZT3 Al beam
Figure 2: The schematic diagram of PZT structure.
Similarly, one has the characters of characteristic param-eters in πth output loop as
π1π(β) = 2,
π2π(β) = β1,
βππ(β) = 0,
π1π(β) + π
2π(β) +
π
β
π=1
βππ(β) = 1.
(9)
Because PZT with small size and light weight does notaffect structureβs characteristics after pasting on the structuresurface, PZT is selected as sensor and actuator. Figure 2indicates the schematic diagram of PZT structure. It is shownthat a pair of PZT1 and PZT2 act as actuator and anotherPZT3 acts as sensor. Additionally, actuators far away PZT3have little function on PZT3. Therefore, (8) can be simplifiedas
π₯π(π + 1) = π
1π(π) π₯π(π) + π
2π(π) π₯π(π β 1)
+ βππ(π) πΉπ(π) = Ξ¦
π
(π) π (π) ,
(10)
where Ξ¦(π) = [π₯π(π) π₯
π(π β 1) πΉ
π(π)]π and π(π) =
[π1π(π) π
2π(π β 1) β
ππ(π)]π.
3. Keep Tracking Control Law
With a proper parameter estimator, the characteristic param-eters can adaptively converge to true values. The generalmethods of parameters identification are least squaremethodand gradient method. Compared with the least squaremethod, gradientmethod has a smaller amount of calculationand it will not cause parameters to diverge with improperforgetting factors.Thus, the gradientmethod [16] will be usedto estimate the characteristic parameters. The form is
οΏ½ΜοΏ½ (π) =
π1Ξ¦ (π β 1) [π₯
π(π) βΞ¦
π
(π β 1) οΏ½ΜοΏ½ (π β 1)]
Ξ¦π
(π β 1)Ξ¦ (π β 1) + π2
+ οΏ½ΜοΏ½ (π β 1) .
(11)
The range ofπ1andπ
2is 0 < π
1< 1, 0 < π
2< 4.The estimates
of π1π(π), π
2π(π), and β
ππ(π) in the characteristic model are
οΏ½ΜοΏ½π
(π) = [οΏ½ΜοΏ½1π(π) οΏ½ΜοΏ½
2π(π) βΜ
ππ(π)].
To ensure the output π¦(π+1) equals the desired referenceinput π(π + 1), the need control input should satisfy thefollowing condition:
π’ (π) =π (π + 1) β οΏ½ΜοΏ½
1π(π) π¦ (π) β οΏ½ΜοΏ½
2π(π) π¦ (π β 1)
βΜππ(π)
. (12)
To avoid the fact of βΜππ(0) = 0, a slack variable π is
introduced, which could guarantee that the control input issmooth and easy to implement. Consider
π’ (π) =π (π + 1) β οΏ½ΜοΏ½
1π(π) π¦ (π) β οΏ½ΜοΏ½
2π(π) π¦ (π β 1)
π + βΜππ(π)
, (13)
where π is a constant. This control law is called characteristicmodel-based keep tracking control (CMKTC) law [16, 17].With the control law, system can perfectly track the referenceinput.
4. PZT Location Optimization
Considering the complexity of traditional optimization algo-rithm,MSE andmaximum amplitude are applied to optimizelocations of PZT whose optimization process is simple.
Firstly beam elements of space structure are dividedinto three categories: underside beam elements π
π’=
{1, 2, . . . , 17, 18}, front and rear beam elements πππ
=
{19, 20, . . . , 33, 34, 36, 37, . . . , 46, 47, 51, 52, . . . , 61, 62}, andconnection beam elements between front and rearππ= {35, 48, 49, 50, 63}. Through beamsβ MSE, the sensitive
beams are obtained. With πth natural frequency, the πthbeamβs MSE [18] can be described as
πΈππ=
ππ
πKπππ
π2
π
, (14)
where ππis modal shape vector with πth natural frequency
which is the πth column of Vnorm. Vnorm = V(βVπMπ)β1is normalized modal matrix, where π is modal matrix. K
πis
πth beam stiffness matrix; ππis πth vibration frequency.
Figures 3β6 show the first four order MSE. Fromthe figures, sensitive beam element sets are obtained:underside sensitive beam element set π
π = {1, 4, 5, 9, 10, 11},
front and rear sensitive beam element set ππ
=
{19, 20, 23, 25, 28, 34, 36, 37}, and connection sensitivebeam element set π
β= {35, 48, 49, 50, 63}. π
π , ππ, and π
β
are the subsets ofππ’, πππ, andπ
π, respectively.
Additionally, it is needed to select the most sensitivebeam element from π
π , ππ, and π
βaccording to maximum
-
Shock and Vibration 5
Table 3: π§-direction maximum amplitude of beams inππ .
Beam number 1 4 5 9 10 11Maximum amplitude (mm) 0.0005 0.0013 0.0015 0.1422 0.3996 0.3996
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 181
36 51 6219
48 49 50 6335
Beam number
Beam number
Beam number
0
1
2
MSE
0
1
2
MSE
0
1
2
MSE
Γ10β3
Γ10β5
Γ10β6
Figure 3: First-order MSE.
48 49 50 6335
36 51 6219
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 181
Beam number
Beam number
Beam number
0
0.005
0.01
MSE
0
0.5
1
MSE
0
1
2
MSE
Γ10β4
Γ10β5
Figure 4: Second-order MSE.
amplitude. Excitation node is selected as node 10 and exci-tation direction is π§. Excitation force π
π= 10 sin(2πππ‘),
and excitation frequency π is selected as the first naturalfrequency 10.03Hz. Through analysing vibration responsecurves of π
π , it is concluded that π§-direction vibration
amplitude is much larger than π₯ and π¦ which is considered.Similarly, analysing vibration response curves ofπ
πandπ
β, it
is shown that π¦-direction vibration amplitude is much largerthan π₯ and π¦. Thus, it is only needed to compare π§-directionof the maximum amplitude for beams inπ
πandπ
β.
From Table 3 which shows the π§-direction maximumvibration amplitude of π
π , 10th or 11th beam element is
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 181
36 51 6219
48 49 50 6335
Beam number
Beam number
Beam number
0
1
2
MSE
0
5
MSE
0
5
MSE
Γ10β3
Γ10β5
Γ10β5
Figure 5: Third-order MSE.
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 181
36 51 6219
48 49 50 6335
Beam number
Beam number
Beam number
0
0.01
0.02
MSE
0
5
MSE
0
0.5
1
MSE
Γ10β5
Γ10β5
Figure 6: Fourth-order MSE.
the most sensitive beam element. In addition, comparedwith the other location on the 10th or 11th beam element,location approaching node 10 has larger vibration amplitudewhich is appropriate location for pasting PZT. Similarly,Table 4 indicates that 19th or 37th beam element is the mostsensitive beam element inπ
πwhile the 19th beam element is
selected. Furthermore, considering that π₯-direction vibrationamplitude nearing node 19 is larger than the other positionon 19th beam element while π¦-direction vibration amplitudeon any location of 19th beam element is nearly the same,so position approaching node 19 on 19th beam element isanother appropriate location for pasting PZT. Lastly, 35th
-
6 Shock and Vibration
Table 4: π¦-direction maximum amplitude of beams inππ.
Beam number 19 20 23 25 28 34 36 37Amplitude (mm) 0.1909 0.0930 0.0423 0.0713 0.0713 0.1909 0.0930 0.1909
Table 5: π¦-direction maximum amplitude of beams inπβ.
Beam number 35 48 49 50 63Maximum amplitude (mm) 0.1909 0.1000 0.0703 0.0135 0.0714
Table 6: Maximum amplitude on 35th beam element.
π (m) π₯ (mm) π¦ (mm) π§ (mm)0 0.00022 0.19 0.000480.2 0.02 0.19 0.0180.4 0.003 0.19 0.0250.6 0.018 0.19 0.0180.8 0.00018 0.19 0.00038
beam element is the most sensitive beam element inπβfrom
Table 5. Since any locationβs π¦-direction vibration amplitudeof 35th beam element has nearly the same value, π§-directionvibration amplitude is considered.Define variable π as the dis-tance between any location on 35th beam element and node19. Table 6 shows that themiddle location with themaximumvibration amplitude is another appropriate location to pastePZT.
In a word, three optimization positions are middle posi-tion on 35th beam element (1st controller CMKTC1), locationapproaching node 19 on 19th beam element (2nd controllerCMKTC2), and location approaching node 10 on 10th beamelement (3rd controller CMKTC3).
5. Numerical Simulation
5.1. Characteristic Model Verification Simulation. To verifythat characteristic model can accurately describe dynamiccharacters of system, with the literature [16], four forms ofcontrol input are selected as follows:
(a) Step signal πΉπ(π) = 10.
(b) Smooth step signal πΉπ(π) = 0.97π’(π β 1) + 0.3.
(c) 10Hz sinusoidal signal πΉπ(π) = 10 sin(20ππΞπ‘).
(d) 10Hz square wave signal πΉπ(π) =
10 sign(sin(20ππΞπ‘)).
Four different control input signals are shown in Figure 7.Control input direction is π§-direction of node 10; vibrationdisplacement output direction is selected as π¦-direction ofnode 19. Output estimation error οΏ½ΜοΏ½(π) can be expressed asοΏ½ΜοΏ½(π) = π₯
π(π) β Ξ¦
π
(π β 1)οΏ½ΜοΏ½(π). Sampling time Ξπ‘ is selectedas 0.001 s.
Figures 8β11 show the output estimation error withdifferent control input. When control input is step signalor smooth step signal, output estimation errorβs magnitudeis 10β7. If control input is sinusoidal signal or square wavesignal, output estimation errorβsmagnitude is 10β5.Therefore,
0
2
4
6
8
10
(a)(b)
(c)(d)
β2
β4
β6
β8
β10
Fi(k)
0.1 0.2 0.3 0.4 0.50Time (s)
Figure 7: Four-control input signal.
Γ10β7
0.2 0.4 0.6 0.8 10
Time (s)
β2
β1.5
β1
β0.5
0
0.5
1
1.5
2
eΜ(k)
Figure 8: Output estimation error with (a).
output of characteristic model can approach the outputof space structure. In addition, characteristic model canaccurately describe space structureβs dynamic characters.
5.2. Distributed PZT Vibration Simulation. Table 7 showsthe related parameters of PZT and Al. With charactersof characteristic parameters (9), select the initial value of
-
Shock and Vibration 7
Table 7: Related parameters of PZT and aluminum.
Thickness (mm) Width (mm) Elastic modulus (Gpa) PSC (C/N)PZT 1 10 8 Γ 1010 5 Γ 10β10
Al 5 12 7 Γ 1010 βRemark 1: PSC: Piezoelectric Strain Constant.
Table 8: Control parameter.
Parameter CMKTC1 CMKTC2 CMKTC3π β0.0005 β0.0005 β0.025
Γ10β7
β1.5
β1
β0.5
0
0.5
1
1.5
eΜ(k)
0.2 0.4 0.6 0.8 10
Time (s)
Figure 9: Output estimation error with (b).
Γ10β5
β6
β4
β2
0
2
4
6
8
eΜ(k)
0.2 0.4 0.6 0.8 10
Time (s)
Figure 10: Output estimation error with (c).
characteristic parameters as οΏ½ΜοΏ½π
(0) = [1.618 β 0.618 0.03].Related parameters of gradient method are π
1= 0.8,
π2
= 0.4. Sampling time Ξπ‘ is selected as 0.001 s. Controlparameters are shown in Table 8. Excitation force π
π=
10 sin(2πππ‘); excitation frequency π is selected as the firstnatural frequency 10.03Hz. Select nodes 19, 22, and 25 asobservation nodes. It should be emphasized that the outputfeedback of CMKTC1 is π¦-direction displacement.
Γ10β5
eΜ(k)
1.5
1
0.5
0
β0.5
β1
β1.5
Time (s)10.80.60.40.20
Figure 11: Output estimation error with (d).
Γ10β4
Γ10β3
Node 19
0.1 0.2 0.3 0.4 0.50
Time (s)
0.1 0.2 0.3 0.4 0.50
Time (s)
0.1 0.2 0.3 0.4 0.50
Time (s)
Without controlCMKTC
β1
0
1
z(m
m)
β0.2
0
0.2
y(m
m)
β5
0
5
x(m
m)
Figure 12: Displacement of node 19.
Figures 12β14 are the vibration displacement curves ofnodes 19, 22, and 25, respectively. It is shown that π¦-directionvibration displacement is much larger than π₯ or π¦ andsuppressing vibration effect of π¦-direction is obvious. WithCMKTC, π¦-direction vibration displacement of nodes 19, 22,and 25 attenuates to 4 Γ 10β3, 4.5 Γ 10β4, and 1.8 Γ 10β4,respectively, which indicates the effectiveness of distributedPZT vibration control. Control input is presented in Figure 15which shows that control voltage is within voltage range ofPZT.
-
8 Shock and Vibration
Table 9: π¦-direction vibration attenuation ratio of nodes 19, 22, and 25 with different amplitude sinusoidal disturbance.
Disturbance amplitude 10 20 30 40 50 60 70 80 90 100Node 19 (%) 96.91 96.52 96.07 95.65 95.18 94.74 94.28 93.82 93.35 92.72Node 22 (%) 96.61 96.71 96.42 96.16 95.90 95.45 95.01 94.54 94.10 93.65Node 25 (%) 96.36 95.27 96.47 93.73 92.00 90.33 88.60 86.90 85.20 83.50
Table 10: π¦-direction vibration attenuation ratio of nodes 19, 22, and 25 with different power-band white noise.
Noise power 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50Node 19 (%) 95.45 94.76 93.82 93.51 92.98 92.50 92.15 91.83 91.37 91.20Node 22 (%) 96.24 96.34 96.21 96.21 96.21 96.21 96.19 96.18 96.16 96.15Node 25 (%) 96.00 95.83 95.73 95.63 95.53 95.43 95.33 95.30 95.23 95.17
Γ10β3
Γ10β4
Node 22
Without controlCMKTC
β1
0
1
x(m
m)
β0.05
0
0.05
y(m
m)
β5
0
5
z(m
m)
0.1 0.2 0.3 0.4 0.50
Time (s)
0.1 0.2 0.3 0.4 0.50
Time (s)
0.1 0.2 0.3 0.4 0.50
Time (s)
Figure 13: Displacement of node 22.
Γ10β4
Γ10β5
Node 25
Without controlCMKTC
0.1 0.2 0.3 0.4 0.50
Time (s)
0.1 0.2 0.3 0.4 0.50
Time (s)
0.1 0.2 0.3 0.4 0.50
Time (s)
β5
0
5
z(m
m)
β0.05
0
0.05
y(m
m)
β5
0
5
x(m
m)
Figure 14: Displacement of node 25.
u1
(V)
0.1 0.2 0.3 0.4 0.50
Time (s)
0.1 0.2 0.3 0.4 0.50
Time (s)
0.1 0.2 0.3 0.4 0.50
Time (s)
β200
0
200
u3
(V)
β200
0
200u2
(V)
β200
0
200
Figure 15: Control input.
To verify the robustness of CMKTC, the matched distur-bance is added. When the sinusoidal matched disturbanceππ= π΄ sin(200π‘) with different amplitude is added, Table 9
shows π¦-direction vibration attenuation ratio of nodes 19,22, and 25. From the table, it is known that node 25is more sensitive than node 19 and node 22. No matterhow, sinusoidal disturbance has small effect on π¦-directionvibration displacement.
Considering that there exists noise in actual system, soband-white noise with different noise power is added. π¦-direction vibration attenuation ratio of nodes 19, 22, and 25is presented in Table 10. The table indicates that vibrationdamping ratio decreases with the increase of noise power.Of course, band-white noise has small effect on systemperformance and CMKTC is with strong robustness.
6. Conclusion
In this paper, a novel distributed PZT control strategybased on characteristic model has been proposed. To avoidthe difficulty in obtaining the mechanical model of spacestructure, an efficient model-free modeling approach, thatis, the so-called characteristic model theory, is adopted.
-
Shock and Vibration 9
Then a PZT position optimization method based on MSEand maximum vibration amplitude has been presented, inwhich the placement optimization of sensors and actuatorsis realized. Due to the established characteristic model, asimple but efficient CMKTC is designed. The power ofthe proposed control strategy is demonstrated on severalnumerical examples.
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper.
Acknowledgment
This work was supported by the National Natural ScienceFoundation of China (no. 61573332).
References
[1] G. L. de Abreu, G. P. de Melo, V. Lopes, and M. J. Brennan,βActive modal damping control of a smart truss structureusing a self-organizing fuzzy controller,β Journal of the BrazilianSociety of Mechanical Sciences and Engineering, vol. 37, no. 2, pp.441β450, 2015.
[2] G. Zhao, J. Wang, and Y. Gu, βOptimal placement of piezoelec-tric active bars in vibration control by topological optimization,βActa Mechanica Sinica, vol. 24, no. 6, pp. 699β708, 2008.
[3] B. Jaishi andW.-X. Ren, βStructural finite elementmodel updat-ing using ambient vibration test results,β Journal of StructuralEngineering, vol. 131, no. 4, pp. 617β628, 2005.
[4] B. Xu and J. S. Jiang, βIntegrated optimization of structureand control for piezoelectric intelligent trusses with uncertainplacement of actuators and sensors,β Computational Mechanics,vol. 33, no. 5, pp. 406β412, 2004.
[5] J.-S. Hwang, A. Kareem, and W.-J. Kim, βEstimation of modalloads using structural response,β Journal of Sound andVibration,vol. 326, no. 3-5, pp. 522β539, 2009.
[6] B. Hazra, A. Sadhu, A. J. Roffel, and S. Narasimhan, βHybridtime-frequency blind source separation towards ambient sys-tem identification of structures,β Computer-Aided Civil andInfrastructure Engineering, vol. 27, no. 5, pp. 314β332, 2012.
[7] M. P. Schoen, R. C. Hoover, S. Chinvorarat, and G. M.Schoen, βSystem identification and robust controller designusing genetic algorithms for flexible space structures,β Journalof Dynamic Systems, Measurement and Control, vol. 131, no. 3,pp. 1β10, 2009.
[8] S. H. Chen, Z. J. Cao, and Z. S. Liu, βAmethod of for the sectionof sensors and actuators location,β Acta Mechanica Sinica, vol.14, no. 4, pp. 353β362, 1998.
[9] W. P. Li and H. Huang, βIntegrated optimization of actuatorplacement and vibration control for piezoelectric adaptivetrusses,β Journal of Sound andVibration, vol. 332, no. 1, pp. 17β32,2013.
[10] Y. Luo, M. Xu, B. Yan, and X. Zhang, βPD control for vibrationattenuation in Hoop truss structure based on a novel piezoelec-tric bending actuator,β Journal of Sound and Vibration, vol. 339,pp. 11β24, 2015.
[11] G. Abreu, M. Brennan, and J. Lopes, βRobust control of anintelligent truss structure,β in Proceedings of the 10th Interna-tional Conference on Recent Advances in Structural Dynamics,Southampton, UK, July 2010.
[12] J. Y. Yang, Z. Q. Liu, X. Cui et al., βExperimental study ofadaptive fuzzy sliding mode control for vibration of a flexiblerectangular plate,β International Journal of Aeronautical andSpace Sciences, vol. 16, no. 1, pp. 28β40, 2015.
[13] J. Lin and Y. B. Zheng, βVibration suppression control of smartpiezoelectric rotating truss structure by parallel neuro-fuzzycontrol with genetic algorithm tuning,β Journal of Sound andVibration, vol. 331, no. 16, pp. 3677β3694, 2012.
[14] S. N. Mahmoodi and M. Ahmadian, βActive vibration controlwith modified positive position feedback,β Journal of DynamicSystems, Measurement and Control, vol. 131, no. 4, 8 pages, 2009.
[15] J. Wilhelm and R. Rajamani, βMethods for multimodal vibra-tion suppression and energy harvesting using piezoelectricactuators,β Journal of Vibration and Acoustics, vol. 131, no. 1, pp.273β283, 2009.
[16] H. X. Wu, J. Hu, and Y. C. Xie, Characteristic Model-BasedIntelligent Adaptive Control, Science and Technology Press ofChina, Beijing, China, 2009.
[17] H. X. Wu, J. Hu, and Y. C. Xie, βCharacteristic model-basedall-coefficient adaptive control method and its applications,βIEEE Transactions on Systems, Man and Cybernetics Part C:Applications and Reviews, vol. 37, no. 2, pp. 213β221, 2007.
[18] A. Preumont, βOn the damping of a piezoelectric truss,β inProceedings of the 21st International Congress of Theoretical andApplied Mechanics, pp. 287β301, Warsaw, Poland, August 2005.
-
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporation http://www.hindawi.com
Journal ofEngineeringVolume 2014
Submit your manuscripts athttp://www.hindawi.com
VLSI Design
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation http://www.hindawi.com
Volume 2014
The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Modelling & Simulation in EngineeringHindawi Publishing Corporation http://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
DistributedSensor Networks
International Journal of