Sei sulla pagina 1di 11

SHEFFIELD HALLAM UNIVERSITY

55-7963 Control of Linear Systems Lab Digital Controllers

Assignment 2
N.A sanuji kalhan 1/31/2013

Student ID: MS 12 9504 78

It is required to design an appropriate control law such that the resulting closed-loop system has an overshoot less than 25% and a settling time less than 8 milliseconds due to a step reference of 1 micrometer. 1. Design a digital PD or PI or PID controller that meets the above design Specifications. R + Controller (Dz) ZOH Hard Disk (Gs) T Y

Digital Controller Design

+ -

Y Controller (Dz) Hard Disk (Gz)

In according to given specification rise time overshoot should be less than 25%. Hence:

1 2

Where, Mp= 0.25 and 0 < 1

0.403 And we know when Mp is going to be decreased, is increased. Therefore to be overshoot less than 25%, = 0.5 is the proposed value for this occurrence. Other specification is settling time, it is given as 8 ms. to meet this specification needs to find proper natural frequency. it can find as following,
4.6

Therefore: = (Still are in continues domain and it will need for SVF designing)

Continuous to Discrete conversion = 1 1 { = 1 1 = 3 107


610 7 2 2 +1 1 2

= where the T=0.0001; =

0.3 +1 1 2

PID controller design

PID Tuning Parameter

Tuned PID Block Parameter

Scope output of PID design (ts =0.006 and Overshoot is almost 10%)

PD controller design PD Block Parameters

Scope output of PD design (ts =0.005 and Overshoot is almost 10% and more speeder than PID)

But PI controller designing for this particular system seems impossible because it is making unexpected oscillation even PI tuning did not work.

PI Tuning

2. Design a digital state feedback controller that meets the design specifications using the

pole placement technique. Once we obtain a state space model for the discrete system, the control system design in the state space setting is straightforward and systematic. We know by now any system can be represented as following way.
+ 1 = + = +

Assuming state variable x(k) is measurable. We can design SF controller according to following structure.
= +

It is substituting to above state space; we can get close loop system as following way.
+ 1 = 1 + 1 ; = 1 + 1 1 = 1 = 1 = 1 =

And there is a new TF for finding DC gain; = 1 1


1

1 + 1 = 1 , = 1

State-space Conversion of Plant C to D


Code: num=[6*10^7]; dem=[1 0 0]; sys1=tf(num,dem); Sp= ss(sys1); Ts= 0.0001; d_sys = c2d(Sp,Ts,'zoh') Results:

Since system is controllable, we can design Digital SF controller considering given conditions. And the matrices A and B*K(where, u=-kx) are both 2by 2 matrices, there will be 2 poles for the system. Then we have to decide where we want the closed-loop poles to be. Suppose it is 2 nd order system and its poles can get considering the characteristic equation of 2 nd order system. = 1 2 We already know values; they are = and = 0.5.

= 575 995.92 || = (575 )0.0001 = 0.944 ; = 995.92 0.0001 = 0.0996 rad = 5.700

Hence:

z = 0.944 [cos(5.700 ) sin(5.700) ] =

0.94 i 0.093 (close loop poles)

State Feedback gain matrix finding using place command


Code: A=[1 0 ;0.0001 1]; B=[ 0.8192 ; 4.096e-005]; C=[ 0 7324]; D=[0]; JJ= [(0.94 + (0.093)*i) (0.94 - (0.093)*i)]; K=place(A,B,JJ) Results

Design a digital state feedback controller


Code: I=[1 0 ;0 1]; A1=A-B*K H(1)=C*((I-A1)^-1)*B J=1/H B1=B*J C1=C-D*K D1=D*J Digitalsp=ss(A1,B1,C1,D1,Ts) Results:

Digital Sate feedback controller

Analyzing through the m-file


step(Digitalsp,0.009)

1.4

Step Response

1.2

The system has successfully followed the design specifications. This system will stable less than 0.008 second (8ms) and overshoot is less than 20%

0.8

Amplitude
0.6 0.4 0.2 0 0

4 Time (sec)

8 x 10

9
-3

Simulink solution when DC-gain given outside from the close-loop state space

In this time system is exactly stable at 8ms and also has overshoot less than 20%

Used Design specifications

2. By now, you have designed the same system using different approaches. Please comment the advantages and disadvantages of these methods. Which one would you recommend? When consider the two methods PID controller and Digital state feedback controller, PID utilization part is very hard and tuning also very complicated when comparing the Digital sate feedback controller. But successfully tuned PID system is giving invaluable outputs. DSF controller no need to tune and just can design through the given design specification. But when DSFC applying to another system need to design it from the beginning considering that particular system specifications. But PID controllers no need to design from beginning just need to tune the controller until meet the system specifications. When consider the above system, both methods has been successfully followed the specification. But PID controller system is stabilized less than 6ms then the DSF controller getting almost 8ms for stabilizing the system. Therefore we can say PID system is giving sensitive immediate response rather than Digital state feedback controller.

Advantages of PID Design Method Kp, Ki, Kd values can be utilized using MATLAB Less calculations Variables can be changed over the time to be adjusted for particular situations

Disadvantages of PID Design Method When Kp, Ki, Kd need to be implemented in real systems (Ex:0.000000043401) therefore High speed computers are needed so that implementation cost is too expensive.

Advantages of Pole Placement Design Method System can be designed accurately as preferred

Disadvantages of Pole Placement Design Method To change the system after the design, is not probable. Need more calculations

I recommended Pole Placement Design Method for hard disk control systems.
Someone says in IEEE: The implementation results of the RPT controller are compared with those of a proportional integral derivative (PID) controller. The results show that the servo system with our RPT controller has much better performance than the PID one has.

Reference: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.139.5512&rep=rep1&type=pdf

Potrebbero piacerti anche