European Journal of Scientific Research ISSN 1450-216X Vol.28 No.4 (2009), pp.583-599 EuroJournals Publishing, Inc. 2009 http://www.eurojournals.com/ejsr.

htm

Optimal Control with Input Shaping for Input Tracking and Vibration Suppression of a Flexible Joint Manipulator
Mohd Ashraf Ahmad Control and Instrumentation Research Group (COINS), Faculty of Electrical and Electronics Engineering, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300, Kuantan, Pahang, Malaysia E-mail: mashraf@ump.edu.my Tel: +609-5492366; Fax: +609-5492377 Raja Mohd Taufika Raja Ismail Control and Instrumentation Research Group (COINS), Faculty of Electrical and Electronics Engineering, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300, Kuantan, Pahang, Malaysia E-mail: rajamohd@ump.edu.my Tel: +609-5492366; Fax: +609-5492377 Mohd Syakirin Ramli Control and Instrumentation Research Group (COINS), Faculty of Electrical and Electronics Engineering, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300, Kuantan, Pahang, Malaysia E-mail: syakirin@ump.edu.my Tel: +609-5492366; Fax: +609-5492377 Abstract This paper presents investigations into the development of optimal control for input tracking and vibration suppression of a flexible joint manipulator. A single-link flexible joint manipulator is considered and the dynamic model of the system is derived using the Euler-Lagrange formulation. To study the effectiveness of the controllers, a linearquadratic regulator (LQR) controller is developed for tip angular position control of a flexible joint manipulator. This is then extended to incorporate input shaper control schemes for vibration reduction of the flexible joint system. The positive zero-vibrationderivative-derivative (ZVDD) and new modified specified negative amplitude zerovibration-derivative-derivative (SNA-ZVDD) input shapers are then designed based on the properties of the system for vibration control. The new SNA-ZVDD is proposed to improve the robustness capability while increasing the speed of the system response. Simulation results of the response of the flexible joint manipulator with the controllers are presented in time and frequency domains. The performances of the LQR with input shaping control schemes are examined in terms of input tracking capability, level of vibration reduction, time response specifications and robustness to parameters uncertainty. A comparative assessment of the positive ZVDD and modified SNA-ZVDD shapers to the hybrid system performance is presented and discussed.

Optimal Control with Input Shaping for Input Tracking and Vibration Suppression of a Flexible Joint Manipulator

584

Keywords: Flexible joint manipulator, vibration control, input shaping, LQR controller.

1. Introduction
It has been recognized that control of robot link based on rigid body dynamic formulation is not enough to deal with stringent condition. As a result, the joint flexibility should be taken into consideration in order to achieved better performance. In some cases, joint flexibility may lead to instability of the system if it is neglected in the control design. Rotary flexible joint is considered as an ideal representation intended to model a flexible joint in a robot manipulator. The control issue of the flexible joint is to design the controller so that link of robot can reach a desired position or track a prescribed trajectory precisely with minimum vibration to the link. In order to achieve these objectives, various methods using different technique have been proposed. Yim [1], Oh and Lee [2] proposed adaptive output-feedback controller based on a backstepping design. This technique is proposed to deal with parametric uncertainty in flexible joint. The relevant work also been done by Ghorbel et al. [3]. Lin and Yuan [4] and Spong et al. [5] introduced non linear control approach using namely feedback linearization technique and the integral manifold technique respectively. A robust control design was reported by Tomei [6] by using simple PD control and Yeon and Park [7] by applying robust H control. Among the proposed techniques, the conventional feedback control design handled by pole placement method and LQR method also have been widely used due to its simplicity implementation. Particularly in LQR method, the values of Q and R matrices are pre-specified to determine optimal feedback control gain via Riccati equation [8]. Input shaping control techniques are mainly develop for vibration suppression and involve developing the control input through consideration of the physical and vibrational properties of the system, so that system vibrations at response modes are reduced. This method does not require any additional sensors or actuators and does not account for changes in the system once the input is developed. Investigations have shown that with the input shaping technique, a system response with delay is obtained. To reduce the delay and thus increase the speed of the response, negative amplitude input shapers have been introduced and investigated in vibration control. By allowing the shaper to contain negative impulses, the shaper duration can be shortened, while satisfying the same robustness constraint. A significant number of negative shapers for vibration control have also been proposed. These include negative unity-magnitude (UM) shaper, specified-negative-amplitude (SNA) shaper, negative zero-vibration (ZV) shaper, negative zero-vibration-derivative (ZVD) shaper and negative zero-vibration-derivative-derivative (ZVDD) shaper [9,10,11]. Comparisons of positive and negative input shapers for vibration control of a single-link flexible manipulator have also been reported [11]. This paper presents investigations into the development of hybrid control schemes for input tracking and vibration control of a flexible joint manipulator system. A flexible joint manipulator system is considered and the dynamic model of the system is derived using the Euler-Lagrange formulation. Hybrid control schemes based on input shaping with LQR controllers are investigated. In this work, input shaping with positive ZVDD input shapers and new modified specified negative amplitude zero-vibration-derivative-derivative (SNA-ZVDD) input shapers are considered. A new modified shaper from the previous SNA input shapers [11] is proposed where more negative impulses are added to improve the robustness of the controller while increasing the speed of the system response. To demonstrate the effectiveness of the proposed control schemes, a LQR controller is developed for control of tip angular position of the flexible joint. This is then extended to incorporate the proposed input shapers for vibration control of flexible joint. Simulation exercises are performed within the flexible joint simulation environment. Performances of the developed controllers are examined in terms of input tracking capability, level of vibration reduction, time response specifications and robustness to errors in vibration frequency. In this case, the robustness of the hybrid control schemes is assessed with up to 30% error tolerance in vibration frequencies. Simulation results

585

Mohd Ashraf Ahmad, Raja Mohd Taufika Raja Ismail and Mohd Syakirin Ramli

in time and frequency domains of the response of the flexible joint to the unshaped input and shaped inputs with positive and modified SNA input shapers are presented. Moreover, a comparative assessment of the effectiveness of the hybrid controllers with positive and negative input shapers in suppressing vibration and maintaining the input tracking capability of the flexible joint is discussed.

The flexible joint manipulator system considered in this work is shown in Figure 1, where is the tip angular position and is the deflection angle of the flexible link. The base of the flexible joint manipulator which determines the tip angular position of the flexible link is driven by servomotor, while the flexible link will response based on base movement. The deflection of link will be determined by the flexibility of the spring as their intrinsic physical characteristics.
Figure 1: Description of the Flexible Joint Manipulator System

2. Gantry Crane System

3. Modelling of the Flexible Joint Manipulator

This section provides a brief description on the modelling of the flexible joint manipulator system, as a basis of a simulation environment for development and assessment of the LQR with input shaping control techniques. The Euler-Lagrange formulation is considered in characterizing the dynamic behaviour of the system. Considering the motion of the flexible joint system on a two-dimensional plane, the potential energy of the spring can be formulated as 1 (1) U = K stiff 2 2 where K stiff is the joint stiffness. The kinetic energies in the system arise from the moving hub and flexible link can be formulated as 1 & 2 + 1 J ( & + &)2 T = J eq (2) link 2 2 where J eq and J link are the equivalent inertia and total link inertia, respectively.

To obtain a closed-form dynamic model of the flexible joint, the energy expressions in (1) and (2) are used to formulate the Lagrangian, that is
L = T U = 1 1 & 2 + 1 J ( & + & ) 2 K stiff 2 J eq link 2 2 2

(3)

Optimal Control with Input Shaping for Input Tracking and Vibration Suppression of a Flexible Joint Manipulator

586

& . Using Let the generalized torque corresponding to the generalized tip angle be output Beq

Lagrangians equation
d dt L L & = output B eq &

(4) (5) (6)

d L L =0 & dt the equation of motion is obtained as below, && J ( && + & &&) = output Beq J eq link && + &&) + K = 0 J (
link stiff

where Beq is the equivalent viscous damping and output the motor, defined as &) K K (V K g K m (8) output = m g t g m Rm where m is the motor efficiency, g is the gearbox efficiency, K t is the motor torque constant, K g is the high gear ratio, K m is the motor back emf constant and Rm is the linkature resistance. The linear model of the uncontrolled system can be represented in a state-space form as shown in equation (9), that is & = Ax + Bu x (9) y = Cx T & & and the matrices A, B and C are given by with the vector x =

(7) is the output torque on the load from

0 0 A = 0 0

0 0 K stiff J eq K stiff (J eq + J arm ) J eq J arm m g K t K g J eq Rm

1 0 2 m g K t K m K g + Beq Rm J eq Rm 2 + Beq Rm m g K t K m K g J eq Rm

0 1 0 0

(10)

B = 0 0

m g K t K g , C = [1 0 0 0] J eq Rm

In equation (9), the input u is the input voltage of the servomotor, V m which determines the flexible joint manipulator base movement. In this study, the values of the parameters are defined as Rm=2.6 , Km=0.00767 V-s/rad, Kt=0.00767 N-m/A, Jlink=0.0035 kg-m2, Jeq= 0.0026 kg-m2, Kg=2.8, Kstiff=1.2485, Beq=0.004 N-m-s/rad, g=0.9 and m=0.69.

4. Linear Quadratic Regulator (LQR) Control Scheme

A more common approach in the control of manipulator systems involves the utilization LQR design [8]. Such an approach is adopted at this stage of the investigation here. Figure 2 illustrates the LQR control structure. In order to design the LQR controller a linear state-space model of the flexible joint manipulator was obtained by linearising the equations of motion of the system. For a linear time invariant (LTI) system

587

Mohd Ashraf Ahmad, Raja Mohd Taufika Raja Ismail and Mohd Syakirin Ramli
Figure 2: Delayed feedback signal controller structure

Desired input

+
N

Flexible Joint Manipulator

Output responses

& = Ax + Bu x (11) the technique involves choosing a control law u = ( x) which stabilizes the origin (i.e., regulates x to zero) while minimizing the quadratic cost function

J = [ x(t ) T Qx(t ) + u (t ) T Ru (t )] dt
0

(12)

where Q = Q T 0 and R = R T > 0 . The term linear-quadratic refers to the linear system dynamics and the quadratic cost function. The matrices Q and R are called the state and control penalty matrices, respectively. If the components of Q are chosen large relative to those of R, then deviations of x from zero will be penalized heavily relative to deviations of u from zero. On the other hand, if the components of R are large relative to those of Q, then control effort will be more costly and the state will not converge to zero as quickly. A famous and somewhat surprising result due to Kalman is that the control law which minimizes J always takes the form u = ( x) = Kx . The optimal regulator for a LTI system with respect to the quadratic cost function above is always a linear control law. With this observation in mind, the closed-loop system takes the form & = ( A BK ) x x (13) and the cost function J takes the form

J = [ x(t ) T Qx(t ) + ( Kx(t )) T R( Kx(t ))] dt

0

= x(t ) T (Q + K T RK ) x(t ) dt
0

(14)

Assuming that the closed-loop system is internally stable, which is a fundamental requirement for any feedback controller, the following theorem allows the computation value of the cost function for a given control gain matrix K.

5. Input Shaping Control Schemes

Input shaping technique is a feed-forward control technique that involves convolving a desired command with a sequence of impulses known as input shaper. The shaped command that results from the convolution is then used to drive the system. Design objectives are to determine the amplitude and time locations of the impulses, so that the shaped command reduces the detrimental effects of system flexibility. These parameters are obtained from the natural frequencies and damping ratios of the system. Thus, vibration reduction of a flexible joint system can be achieved with the input shaping technique. Figure 3 illustrates the input shaping process. Several techniques have been investigated to obtain an efficient input shaper for a particular system. A brief description and derivation of the control technique is presented in this section.

Optimal Control with Input Shaping for Input Tracking and Vibration Suppression of a Flexible Joint Manipulator
Figure 3: Illustration of input shaping technique.

588

Amplitude

Time Unshaped Input

A2

Amplitude

A1

Time Shaped input

Input Shaper

Generally, a vibratory system of any order can be modelled as a superposition of second order systems each with a transfer function
G( s) =

2 s 2 + 2s + 2

(15)

where is the natural frequency of the vibratory system and is the damping ratio of the system. Thus, the response of the system in time domain can be obtained as A y (t ) = e ( t t 0 ) sin[ 1 2 (t t 0 )] (16) 2 1 where A and t 0 are the amplitude and the time location of the impulse respectively. The response to a sequence of impulses can be obtained by superposition of the impulse responses. Thus, for N impulses, with d = 1 2 , the impulse response can be expressed as y (t ) = M sin( d t + ) where
2 2

(17)

Ai N N e (t t0 ) , i = d t i M = Bi cos i + Bi sin i , Bi = i =1 i =1 1 2 and Ai and ti are the amplitudes and time locations of the impulses. The residual single mode vibration amplitude of the impulse response is obtained at the time of the last impulse, t N as
V = V12 + V 22

(18)
e n ( t N t i ) cos( d t i ) ; V 2 =

where
V1 =
i =1 N

Ai n 1
2

1 To achieve zero vibration after the last impulse, it is required that both V1 and V2 in Equation (18) are independently zero. This is known as the zero residual vibration constraints. In order to ensure that the shaped command input produces the same rigid body motion as the unshaped reference command, it is required that the sum of amplitudes of the impulses is unity. This yields the unity amplitude summation constraint as
i =1 2

Ai n

e n ( t N ti ) sin( d t i )

A
i =1

=1

(19)

In order to avoid response delay, time optimality constraint is utilised. The first impulse is selected at time t1 = 0 and the last impulse must be at the minimum, i.e. min(t N ) . The robustness of the input shaper to errors in natural frequencies of the system can be increased by taking the derivatives of V1 and V2 to zero. Setting the derivatives to zero is equivalent to producing small changes in vibration corresponding to the frequency changes. The level of robustness can further be increased by increasing

589

Mohd Ashraf Ahmad, Raja Mohd Taufika Raja Ismail and Mohd Syakirin Ramli

the order of derivatives of V1 and V2 and set them to zero. Thus, the robustness constraints can be obtained as d iV2 d iV1 = 0 ; =0 (20) i i d n d n Both the positive and modified SNA input shapers are designed by considering the constraints equations. The following section will further discuss the design of the positive and modified SNA input shapers.
5.1. Positive Input Shaper

The positive input shapers have been used in most input shaping schemes. The requirement of positive amplitude for the impulses is to avoid the problem of large amplitude impulses. In this case, each individual impulse must be less than one to satisfy the unity magnitude constraint. In order to increase the robustness of the input shaper to errors in natural frequencies, the positive ZVDD input shaper is designed by setting the second derivatives of V1 and V2 in Equation (18) to zero. Simplifying
2 d 2V i d n yields

(21) 2 2 d n d n i =1 i =1 The positive ZVDD input shaper, i.e. four-impulse sequence is obtained by setting Equations (18) and (21) to zero and solving with the other constraint equations. Hence, a four-impulse sequence can be obtained with the parameters as 2 3 t1 = 0 , t 2 = , t3 = , t4 =

d 2V1

= Ai t i2 e n ( t N ti ) sin( d t i );

d 2V 2

= Ai t i2 e n ( t N ti ) cos( d t i )

d d d 1 3K , A2 = A1 = 2 3 1 + 3K + 3 K + K 1 + 3 K + 3K 2 + K 3 3K 2 K3 A3 = A = 4 1 + 3K + 3 K 2 + K 3 , 1 + 3 K + 3K 2 + K 3
1 2

(22)

where

K =e d = n 1 2 n and representing the natural frequency and damping ratio respectively. For the impulses, t j and A j are the time location and amplitude of impulse j respectively.
5.2. Modified SNA Input Shapers

Input shaping techniques based on positive input shaper have been proved to be able to reduce vibration of a system. In order to achieve higher robustness, the duration of the shaper is increased and thus, increases the delay in the system response. By allowing the shaper to contain negative impulses, the shaper duration can be shortened, while satisfying the same robustness constraint. To include negative impulses in a shaper requires the impulse amplitudes to switch between 1 and 1 as Ai = (1) i +1 ; i = 1,2, K , n (23) The constraint in Equation (23) yields useful shapers as they can be used with a wide variety of inputs. For a unity magnitude (UM) negative zero-vibration (ZV) shaper, i.e. the magnitude of each impulse is |1|, the shaper duration is one-third of the vibration period of an undamped system, while the shaper duration for the positive shaper is half of the vibration period. However, the increase in the speed of system response achieved using the SNA input shapers is at the expense of some tradeoffs and

Optimal Control with Input Shaping for Input Tracking and Vibration Suppression of a Flexible Joint Manipulator

590

penalties. The shapers containing negative impulses have tendency to excite unmodeled high modes and they are slightly less robust as compared to the positive shapers. Besides, negative input shapers require more actuator effort than the positive shapers due to high changes in the set-point command at each new impulse time location. To overcome the disadvantages, a modified SNA input shaper is introduced, whose negative amplitudes can be set to any value at the centre between each normal impulse sequences. In this work, the previous SNA input shaper [11] is modified by locating the negative amplitudes at the centre between each positive impulse sequences with even number of total impulses. This will result the shaper duration as one-fourth of the vibration period of an undamped system as shown in Figure 4. The modified SNA-ZVDD shaper is proposed and applied in this work to enhance the robustness capability of the controller while increasing the speed of the system response. By considering the form of modified SNA-ZVDD shaper shown in Figure 4, the amplitude summation constraints equation can be obtained as 2a + 2c 2b 2d = 1 (24) The values of a, b, c and d can be set to any value that satisfy the constraint in (24). However, the suggested values of a, b, c and d are less than |1| to avoid the increase of the actuator effort.
Figure 4: Modified SNA-ZVDD shaper.
c a 0.5t2 b 1.5t2 d 2.5t2 d c a 3.5t2 b

t2

t3

t4

6. Implementation and Result

In this section, the proposed control schemes are implemented and tested within the simulation environment of the flexible joint manipulator and the corresponding results are presented. In this work, positive ZVDD and modified SNA-ZVDD are investigated as the input shaping control schemes. The tip angle position of the flexible joint is required to follow a trajectory within the range of 50 as shown in Figure 5. System responses namely the tip angular position, angular velocity and deflection angle of the flexible joint are observed. To investigate the vibration of the system in the frequency domain, power spectral density (PSD) of the response at the deflection angle is obtained. The performances of the hybrid controllers are assessed in terms of input tracking and vibration suppression in comparison to the LQR control. Moreover, robustness of the controllers to variations in vibration frequency is also investigated. In this case, 30% error tolerance in vibration frequency is considered.

591

Mohd Ashraf Ahmad, Raja Mohd Taufika Raja Ismail and Mohd Syakirin Ramli
Figure 5: The Trajectory Reference Input
60

40 Desired input (deg)

20

-20

-40

-60 0

10

15 Time (s)

20

25

30

6.1. LQR Control

In this investigation, the responses of the flexible joint manipulator system to the unshaped trajectory reference input were analyzed in time-domain and frequency domain (spectral density) as shown in Figure 6. These results were considered as the system response to the unshaped input under tracking capability and will be used to evaluate the performance of the input shaping techniques. The steadystate tip angular position trajectory of +50 for the flexible joint was achieved within the rise and settling times and overshoot of 0.489 s, 1.235 s and 4.08% respectively. It is noted that the flexible joint reaches the required position from +50 to 50 within 2.3 s, with low overshoot. However, a noticeable amount of deflection angle occurs during movement of the tip angular. It is noted from the deflection angle response with a maximum residual of 7.4. Moreover, from the PSD of the deflection angle response, the vibration frequency is dominated by a single mode, which is obtained as 6 Hz with magnitude of 0.8668 deg2/Hz respectively. The closed loop parameters with the LQR control will subsequently be used to design and evaluate the performance of hybrid controllers with positive ZVDD and SNA-ZVDD shapers.
Figure 6: Response of the flexible joint manipulator with LQR controller.
60

40 Tip angular position (deg)

20

-20

-40

-60 0

10

15 Time (s)

20

25

30

(a) Tip angular position.

Optimal Control with Input Shaping for Input Tracking and Vibration Suppression of a Flexible Joint Manipulator
200 150 Angular velocity (deg/s) 100 50 0 -50 -100 -150 -200 0

592

10

15 Time (s)

20

25

30

(b) Angular velocity.

8 6 4 2 0 -2 -4 -6 -8 0

Deflection angle (deg)

10

15 Time (s)

20

25

30

(c) Deflection angle.

Magnitude ((deg)*(deg)/Hz)

10

10

-5

10

-10

10

-15

10

20

30 40 Frequency (Hz)

50

60

(d) PSD of deflection angle.

6.2. Hybrid Control

Figure 7 shows a block diagram of the proposed hybrid control scheme where the LQR is combined with the input shaping control schemes. The positive ZVDD and modified SNA-ZVDD shapers were designed based on the dynamic behaviour of the closed-loop system obtained using only the LQR controller. As demonstrated in the previous section, the natural frequency of the deflection angle was obtained at 6 Hz for the single mode of vibration. With exact natural frequencies, the time locations

593

Mohd Ashraf Ahmad, Raja Mohd Taufika Raja Ismail and Mohd Syakirin Ramli

and amplitudes of the impulses for positive ZVDD shaper were obtained by solving Equation (23). Moreover, the amplitudes of the modified SNA-ZVDD shaper were deduced as [0.3 0.1 0.5 0.2 0.5 0.2 0.3 0.1] and the time locations of the impulses were chosen at the half of the time locations of positive ZVDD shaper as shown in Figure 4. For evaluation of robustness, input shapers with erroneous natural frequencies were also evaluated. With 30% error in natural frequency, the system vibration was considered at 7.8 Hz for the single mode of vibration. Similarly, the amplitudes and time locations of the input shapers with 30% erroneous natural frequencies for both the positive and modified SNA-ZVDD shapers were calculated.
Figure 7: Block diagram of the hybrid control schemes configuration.
LQR Controller Desired input Input shaper Shaped input + Flexible joint manipulator Output responses

For digital implementation of the input shaper, locations of the impulses were selected at the nearest sampling time. The developed input shaper was then used to pre-process the input reference shown in Figure 5. Figure 8 shows the shaped inputs using both the positive and modified SNA-ZVDD shapers with exact natural frequencies. It is noted that the shaped input with the modified SNA shaper is not as smooth as compared to the positive shaper. This is due to higher number of switching of the actuator.
Figure 8: Shaped inputs with exact natural frequencies using positive ZVDD and modified SNA-ZVDD shapers.
60 Positive ZVDD Modified SNA ZVDD 60 Desired input (deg) 20 40 0 20

40

-20

-40

0 0

0.5

-60 0

10

15 Time (s)

20

25

30

Figure 9 shows the system responses of the flexible joint using the hybrid controllers with exact natural frequencies. Table 1 summarises the levels of vibration reduction of the system responses at the single mode in comparison to the LQR control. It is noted that the proposed hybrid controllers are capable of reducing the system vibration while maintaining the input tracking performance of the tip angular position. Similar tip angular position and tip angular velocity responses were observed as compared to the LQR controller. Moreover, a significant amount of vibration reduction was

Optimal Control with Input Shaping for Input Tracking and Vibration Suppression of a Flexible Joint Manipulator

594

demonstrated at the deflection angle of the flexible joint with both control schemes. With the positive ZVDD and modified SNA-ZVDD shapers, the maximum deflection angles were obtained at 2.6 and 3.5 respectively. These are three-fold and two-fold improvements as compared to LQR controller. This is also evidenced in the PSD of the deflection angle residual that shows lower magnitude at the resonance mode. The corresponding rise time, settling time and overshoot of the tip angular response using LQR controller with positive and modified SNA ZVDD shapers with exact natural frequencies is depicted in Table 1. The simulation results show that the tip angular position reaches the required trajectory position of +50 within the settling times of 1.329 s and 1.277 s with positive ZVDD and modified SNA-ZVDD respectively. It is noted with the input shaping controller, a slower settling time as compared to the LQR controller was achieved. To examine the robustness of the hybrid controllers, the shapers with 30% error in vibration frequency were designed and implemented to the flexible joint manipulator system. Figure 10 shows the response of the flexible joint with the hybrid controllers with erroneous natural frequency. Table 1 summarises the levels of vibration reduction with erroneous natural frequency in comparison to the LQR controller. The time response specifications of the tip angular position with error in natural frequency are also summarised in Table 1. Similar to the case with exact frequency, the proposed hybrid controllers are capable of reducing the system vibration while maintaining the input tracking performance of the tip angular position. Moreover, the vibration of the system was considerable reduced as compared to the response with LQR controller. The simulation results show that performance of the hybrid controller with positive ZVDD control scheme is better than SNA-ZVDD scheme in vibration suppression of the flexible joint for the exact natural frequency case. This is further evidenced in Figure 11 that demonstrates the level of deflection angle reduction of the flexible joint with the hybrid controllers as compared to the LQR controller. It is noted that higher deflection angle reduction is achieved with positive ZVDD at the exact natural frequency of 6 Hz. Almost less than two-fold improvement in the vibration reduction was observed as compared to SNA-ZVDD. Comparisons of the tip angular position responses show that the hybrid controller with SNA-ZVDD shaper is faster than the case using the positive ZVDD shaper. The result reveals that the speed of the system responses can be improved by using negative impulses input shaper. Comparison of the results shown in Figure 11 also reveals that both hybrid controllers with the positive and SNA input shapers can successfully handle errors in natural frequencies. It shows that, almost three-fold improvement was observed with SNA-ZVDD as compared to positive ZVDD deflection angle reduction of the flexible joint. As positive ZVDD performs better than SNA-ZVDD with exact natural frequency, the results demonstrate that, the modified SNA-ZVDD is capable of improving the robustness of the controller to uncertainty in vibration frequency. Comparisons of the tip angular position response with the hybrid controllers show a similar pattern as the case with exact natural frequency. With the new proposed SNA shaper, it is shown that the robustness of the controller can be improved while increasing the speed of the response. The work thus developed and reported in this paper forms the basis of design and development of hybrid control schemes for input tracking and vibration suppression of flexible joint and flexible link manipulator systems and can be extended to and adopted in practical applications.

595
Table 1:

Mohd Ashraf Ahmad, Raja Mohd Taufika Raja Ismail and Mohd Syakirin Ramli
Level of deflection angle reduction and specifications of the tip angular position response for the hybrid control schemes
Types of shaper (ZVDD) Positive 29.14 Positive 69.42 Attenuation (dB) of deflection angle in frequency domain 53.68 0.517 25.15 0.506 Specifications of tip angular position response Rise time (s) 0.515 1.277 0.505 1.275 Settling time (s) 1.329 2.76 1.313 2.88 Overshoot (%) 2.88 3.14

Frequency Exact Modified SNA Error Modified SNA

Figure 9: Response of the flexible joint manipulator with hybrid controllers with exact natural frequencies.
60 Positive ZVDD Modified SNA ZVDD -30

40 Tip angular position (deg)

20

-35

40

-40 10.610.6510.710.75

-20

35

-40

30 0.45 0.50 0.55 0.60 5 10 15 Time (s) 20 25 30

-60 0

(a) Tip angular position

200 150 Angular velocity (deg/s) 100 50 0 -50 -100 -150 -200 0 100 50 0 20 5 10 15 Time (s) 21 20 25 22 30 Positive ZVDD Modified SNA ZVDD

(b) Angular velocity.

Optimal Control with Input Shaping for Input Tracking and Vibration Suppression of a Flexible Joint Manipulator
8 6 4 2 0 -2 -4 -6 -8 0 -2 0 5 0.5 10 1 15 Time (s) 20 1.5 25 2 30 2 0 Positive ZVDD Modified SNA ZVDD

596

Deflection angle (deg)

(c) Deflection angle.

Positive ZVDD Modified SNA ZVDD Magnitude ((deg)*(deg)/Hz) 10

0

10

-5

10

-10

10

-15

10

20

30 40 Frequency (Hz)

50

60

(d) PSD of deflection angle.

Figure 10: Response of the flexible joint manipulator with hybrid controllers with erroneous natural frequencies.
60 Positive ZVDD Modified SNA ZVDD -30

40 Tip angular position (deg)

20

-35

40 35

-40 10.5510.6010.6510.70

-20

-40 30 0.45 0.50 0.55 0.60 -60 0 5 10 15 Time (s) 20 25 30

(a) Tip angular position.

597

Mohd Ashraf Ahmad, Raja Mohd Taufika Raja Ismail and Mohd Syakirin Ramli
200 150 Angular velocity (deg/s) 100 50 0 -50 100 -100 -150 -200 0 50 0 0 5 10 5 10 20 15 25 20 30 Positive ZVDD Modified SNA ZVDD

15 Time (s)

(b) Angular velocity.

8 6 4 2 0 -2 -4 -6 -8 0 -2 0 5 0.5 10 1 15 Time (s) 1.5 20 2 25 30 2 0 Positive ZVDD Modified SNA ZVDD

Deflection angle (deg)

(c) Deflection angle.

Positive ZVDD Modified SNA ZVDD Magnitude ((deg)*(deg)/Hz) 10

0

10

-5

10

-10

10

-15

10

20

30 40 Frequency (Hz)

50

60

(d) PSD of deflection angle.

Optimal Control with Input Shaping for Input Tracking and Vibration Suppression of a Flexible Joint Manipulator

598

Figure 11: Level of Deflection Angle Reduction with Exact and Erroneous Natural Frequencies with Hybrid Controllers.
80 Level of deflection angle reduction (dB) 70 60 50 40 30 20 10 0 Exact Natural frequencies Error
Positive ZVDD Modified SNA ZVDD

7. Conclusion
The development of hybrid control schemes based on LQR control with positive and negative input shapers for input tracking and vibration suppression of a flexible joint manipulator has been presented. The proposed control schemes have been implemented and tested within simulation environment of a flexible joint manipulator system derived using the Euler-Lagrange formulation. The performances of the control schemes have been evaluated in terms of input tracking capability, level of vibration reduction, time response specifications and robustness. Acceptable performance in input tracking control and vibration suppression has been achieved with both control strategies. Moreover, a significant reduction in the system vibration has been achieved with the hybrid controllers regardless of the polarities of the shapers. For exact natural frequency case, a comparative assessment of the hybrid control schemes has shown that the LQR controller with positive ZVDD shaper provides higher level of vibration reduction of the flexible joint as compared to the LQR controller with SNA-ZVDD shaper. However, by using the LQR controller with modified SNA-ZVDD, the robustness of the controller can be improved with slightly increased in the speed of response.

599

Mohd Ashraf Ahmad, Raja Mohd Taufika Raja Ismail and Mohd Syakirin Ramli Yim, W., 2001. Adaptive Control of a Flexible Joint Manipulator. Proc. 2001 IEEE, International Robotics & Automation, Seoul, Korea, pp. 34413446. Oh, J. H. and Lee, J. S., 1997. Control of Flexible Joint Robot System by Backstepping Design Approach. Proc. of IEEE International Conference on Robotics & Automation, pp. 34353440. Ghorbel, F., Hung, J.Y. and Spong, M.W., 1989. Adaptive Control of Flexible Joint Manipulators. Control Systems Magazine, Vol. 9, pp. 9-13. Lin, L.C. and Yuan, K., 2007. Control of Flexible Joint Robots via External Linearization Approach. Journal of Robotic Systems, Vol. 1 (1), pp. 1-22. Spong, M. W., Khorasani, K. and Kokotovic, P. V., 1987. An Integral Manifold Approach to the Feedback Control of Flexible Joint Robots. IEEE Journal of Robotics and Automation, Vol. RA-3, No.4, pp. 291-300. Tomei, P., 1991. A Simple PD Controller for Robots with Elastic Joints. IEEE Trans. on Automatic Control, Vol. 36, No. 10, pp. 1208-1213. Yeon, J. S. and Park, J. H., 2008. Practical Robust Control for Flexible Joint Robot Manipulators. Proc. 2008 IEEE International Conference on Robotic and Automation, Pasadena, CA, USA, pp. 33773382. Ogata, K., 2002. Modern Control Engineering, Prentice Hall. Singhose, W., Singer, N. C. and Seering, W. P., 1994. Design and Implementation of Timeoptimal Negative Input Shapers. Proc. of International Mechanical Engineering Congress and Exposition, Chicago, pp. 151-157. Singhose, W. and Mills, B. W., 1999. Command Generation Using Specified-negativeamplitude Input Shapers. Proc. of the American Control Conference, San Diego, California, pp. 61-65. Mohamed, Z., Chee, A.K., Mohd Hashim, A. W. I., Tokhi, M. O., Amin, S. H. M. and Mamat, R., 2006. Techniques for Vibration Control of a Flexible Manipulator. Robotica 24, pp. 499511.

References
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11]