You are on page 1of 14

Chapter 1

The control of nonlinear systems has been a significant research topic and many approaches have
been proposed. In most of the process industries controlling of level, flow, temperature and
pressure is an exigent one. They may be classified as linear and non-linear processes based on the
plant dynamics. Control of industrial processes is a challenging task for several reasons due to
their nonlinear behavior, uncertain and time varying parameters, constraints on manipulated
variable, interaction between manipulated and controlled variables, unmeasured and frequent
disturbances, dead time on input and measurements. The control of liquid level in tanks and flow
between the tanks is a basic crisis in process industries. In level control process, the tank systems
like cylindrical, cubical are linear one, but that type of tanks does not provides a complete drainage.
For complete drainage of fluids, a conical tank is used in some of the process industries, where its
nonlinearity might be at the bottom only in the case of conical bottom tank. The drainage
efficiency can be improved further if the tank is fully conical in shape. In many processes such as
distillation columns, evaporators, re-boilers and mixing tanks, the particular level of liquid in the
vessel is of great significance in process operation. A level that is too high may upset reaction
equilibrium, cause damage to equipment or result in spillage of valuable or hazardous material. If
the level is too low it may have bad consequences for the sequential operations. So control of liquid
level is an important and frequent task in process industries. Level of liquid is desired to
maintain at a constant value. This is achieved by controlling the input flow. The control variable
is the level in a tank and the manipulated variable is the inflow to the tank. Conical tanks find wide
applications in process industries, namely hydrometallurgical industries, food process industries,
concrete mixing industries and wastewater treatment industries.

The control of fluid level in tanks is an essential issues in process industries. The nonlinear system
show numerous testing control issues because of their nonlinear vibrant deeds and time changing
constraint. The conical tank shows its nonlinearity because of its shape. Design a controller for a
nonlinear process is perplexing and excessively hard to implement it. The principle assignment of
the controller configuration is to accomplish the preferred working conditions and to design the
controller to attain its optimum execution performance. There is a necessitate to control a Level
due to the fact that if the level is excessively high may annoy its reaction equilibrium of the entire

methodology which may cause harm to equipment, or bring out spillage of profitable or risky
material from the process. If the event that the level is excessively low, it may have terrible results
for the sequential operation completed by the process. Henceforth, control of fluid level is a
paramount and common chore in the methodology of process industries. Nonlinear models are
utilized where precision over a more extensive range of operation is obliged where they can be
specifically incorporated into control algorithms. Due to the innate nonlinearity most of the
chemical process industries are in need of innovative control techniques. The nonlinearity is
because of its change in shape. Their shape assures optimal rousing and mixing of ingredients and
provides a fast and hygienic cleaning. The flexibility of the digital computer, digital control
algorithms need not be restricted to discrete versions of analog designs. In particular, it is possible
to formulate controllers that, under ideal conditions, will produce desired closed loop response.
Basically, digital controllers are focused based on its process models, which will have very few
special cases, the design start with the determination of some desired closed loop properties.
Distinctive controllers of diverse complexities will result depending upon the criterion and the
form of the process model.

The system used is a conical tank and is highly nonlinear due to the variation in area of cross
section. The controlling variable is inflow of the tank. The controlled variable is level of the
conical tank. Level sensor is used to sense the level in the process tank and fed into the signal
conditioning unit and the required signal is used for further processing.

The level process station used to perform the experiments and to collect the data. One of the
computer used as a controller. It consists of the software which is used to control the level process
station. The Fig 2 consists of a process tank, reservoir tank, control valve, I to P converter, level
sensor and pneumatic signals from the compressor. When the setup is switched on, level sensor
senses the actual level, initially the signal is converted to current signal in the range between 4 to
20mA. This signal is then given to the computer through data acquisition cord. Based on the
controller parameters and the set point value, the computer will take consequent control action
and the signal is sent to the I/P converter.

Then the signal is converted to pressure signal using I to P converter and the pressure signal acts
on a control valve which controls the inlet flow of water in to the tank. Capacitive type level sensor
is used to senses the level from the process and converts into electrical signal. Then the electrical
signal is fed to the I/V converter which in turn produces corresponding voltage signal to the
computer. The actual water level storage tank sensed by the level transmitter is feedback to the
level controller & compared with a desired level to produce the required control action that will
position the level control as needed to maintain the desired level. Now the controller decides the
control action & it is given to the V/I converter and then to I/P converter. The final control element
(pneumatic control valve) is now controlled by the resulting air pressure. This in turn control the
inflow to the conical tank & the level is maintained. The tank is made up of stainless steel body
and is mounted over a stand vertically. Water enters the tank from the top and leaves the bottom
to the storage tank. The System specifications of the tank are as follows chapter.

Figure 1: Block diagram of process

III. Mathematical Modelling
The mathematical model of the conical tank is deciding by considering two assumptions

(i) By taking level as the control variable and

(ii) Inflow to the tank as the manipulated variable. This is accomplished by controlling the
input flow into the tank. The Figure 1 shows the schematic diagram of the conical tank

The Conical tank as shown in Fig. 1 represents a nonlinear process which can be described by the
following equation


𝑞𝑂 = 𝐶𝐷 ℎ

Fig 2: Conical Tank

Unsteady state mass balance (Assume R to be linear)

𝜌𝑞𝐼 + 𝜌𝑞𝑂 = 𝜌 𝑑𝑡 (1)

Volume at level h

ℎ𝜋𝑑 2 𝜋ℎ 𝐷ℎ 2
𝑣= = ( ) = 𝐾 ′′ ℎ3
12 12 𝐻
ℎ 𝑑(ℎ3 )
𝑞𝐼 + = 𝐾 ′′
𝑅 𝑑𝑡

At steady state the above equation becomes
ℎ𝑠 𝑑(ℎ𝑠3 )
𝑞𝐼𝑠 + = 𝐾 ′′
𝑅 𝑑𝑡
Introducing the non-linearity in resistance and apply taylor’s expansion
𝐶 𝑑(ℎ−ℎ𝑠 )
(𝑞𝐼 − 𝑞𝐼𝑠 ) + 2 𝐷ℎ (ℎ − ℎ𝑠 ) = 3ℎ𝑠2 𝐾 ′′ (2)

Introducing deviation variables:

𝐶 𝑑𝐻
𝑄𝐼 + 2 𝐷ℎ 𝐻 = 3ℎ𝑠2 𝐾 ′′ (3)

Taking Laplace transform and rearranging the Transfer Function, we obtain:

𝐻(𝑠) 𝐶𝑑 𝐾
= =
𝑄𝐼 (𝑠) ℎ2 𝜋𝐷2 2√ℎ 𝜏𝑠 + 1
( 𝑠 2 ) ( 𝐶 𝑠) 𝑠 + 1
4𝐻 𝑑

Table 1: Transfer models of Conical tank at different operating point and IMC PI values

Transfer Models Kp Ti Transfer Models Kp Ti

0.604724 0.378986
M1 = 364.0369𝑠+1 3.307292 364.0369 M7 = 57.0362𝑠+1 5.277247 57.0362
0.580019 0.326023
M2 = 293.3955𝑠+1 3.44816 293.3955 M8 = 34.0733𝑠+1 6.134529 34.0733
0.549444 0.299588
M3 = 229.6938𝑠+1 3.64004 229.6938 M9 = 20.03875𝑠+1 6.675824 20.03875
0.505294 0.24537
M4 = 171.1019𝑠+1 3.958091 171.1019 M10 = 9.231879𝑠+1 8.150943 9.231879
0.437148 0.185185
M5 = 116.959𝑠+1 4.575114 116.959 M11 = 3.096647𝑠+1 10.8 3.096647
0.415051 0.11111
M6 = 85.02027𝑠+1 4.818691 85.02027 M12 = 0.464497𝑠+1 18 0.464497

Table 2: Minimized model by using gap metric (Metric value 0.1)

Transfer Models Kp Ti
M3 = 229.6938𝑠+1 3.64004 229.6938
M7 = 57.0362𝑠+1 5.277247 57.0362
M10 = 9.231879𝑠+1 8.150943 9.231879

Table 3: Weighting value based on gap metric for particular operating point

Operating Range IMC PID 1 IMC PID 2 IMC PID 3

0-20 cm 0.3874 0.3254 0.2872

20-40 cm 0.3169 0.3773 0.3058

40-60cm 0.2905 0.3176 0.3919

Gap Metric Values

ga = 0.0010 0.0819 0.1551 0.2332 0.3141 0.3805 0.4400 0.4683 0.4605

gb = 0.0819 0.0010 0.0780 0.1600 0.2458 0.3229 0.3932 0.4361 0.4312

gc = 0.1551 0.0780 0.0010 0.0849 0.1746 0.2624 0.3444 0.4019 0.4136

gd = 0.2332 0.1600 0.0849 0.0010 0.0927 0.1922 0.2849 0.3600 0.3922

ge = 0.3141 0.2458 0.1746 0.0927 0.0010 0.1132 0.2146 0.3073 0.3658

gf = 0.3805 0.3229 0.2624 0.1922 0.1132 0.0010 0.1102 0.2215 0.3054

gg = 0.4400 0.3932 0.3444 0.2849 0.2146 0.1102 0.0010 0.1259 0.2400

gh = 0.4683 0.4361 0.4019 0.3600 0.3073 0.2215 0.1259 0.0010 0.1395

gi = 0.4605 0.4312 0.4136 0.3922 0.3658 0.3054 0.2400 0.1395 0.0010

Chapetr 2

I. Design of digital control systems with dead beat response

So far we have discussed the design methods which are extensions of continuous time design
We will now deal with the dead beat response design of digital control system. We must distinguish
between the designs of deadbeat response for a digital control system, where all the components
are subject to only digital data, and a sampled data control system, where both continuous and
discrete components are present. An all digital control system is shown in Figure 3.

Fig 3: An all digital control system

Deadbeat response design when the system poles and zeros are
inside the unit circle
Design criteria:
1) The system must have a zero steady state error at sampling instants.
2) The time to reach final output must be finite and minimum.
3) The controller should be physically realizable, i.e., it should be causal. We can write from

𝐶(𝑧) 𝐷𝑐(𝑧)𝐺𝑝(𝑧)
M(z) = =
𝑅(𝑧) 1+𝐷𝑐(𝑧)𝐺𝑝(𝑧)


N: positive integer
A(z): polynomial in z-1 with no zeros at z = 1.

For unit step signal A(z) = 1 and N = 1.

For unit ramp signal A(z) = T z-1 and N = 2.
To achieve zero steady state error

Since A(z) does not contain any zero at z = 1, necessary condition for zero steady state error
is that 1 - M(z) should contain (1 - z-1)N as a factor, i.e.,

F(z) is a polynomial in z-1.
Q(z) is a polynomial in z.

Substituting M(z) in the expression of E(z), E(z) = A(z)F(z). Since A(z) and F(z) are both
polynomials of z-1, E(z) will have a finite number of terms in the power series in the inverse
power of z, i.e., the error will go to zero in a finite number of sampling periods.

II. Dead Beat Algorithm

Simulation is performed using MATLAB for the Deadbeat Algorithm to validate theperformance
Servo and regulatory response is determined for the setpoint of 30cm and 40cm to control the
level of the tank using deadbeat controller. The simulation is carried out by taking 60% and 80%
as the nominal value . It provides minimum settling time and very high stable output .In figure 4
and 5 gives the closed loop response of a deadbeat controller having the setpoint of 30 cm and
40cm correspondingly of the tank level .Figure 6 and 7 shows the Closed loop response of a
deadbeat controller having setpoint 30cm with the setpoint change of 32cm, and Closed loop
response of a deadbeat controller having setpoint 40cm with the setpoint change of 42cm, Figure
8 provides Servo and Regulatory response of Closed loop system with deadbeat controller
having setpoint 40cm , Figure 8 shows the Servo and Regulatory response of Closed loop system
with deadbeat controller having setpoint 40cm. Figure 9and 10 gives he output of Regulatory
response of Closed loop system with deadbeat controller having setpoint 40cm and 30 cm

III. Simulation Results

Fig 4: Simulation diagram for conical system

Fig. 5 Closed loop response of a deadbeat controller having setpoint 30cm

Fig 6: Simulation diagram for conical system with Deadbeat controller

Fig. 7 Closed loop response of a deadbeat controller having set point 30cm with the set point
change of 32cm

Fig. 8 Servo and Regulatory response of Closed loop system with deadbeat controller having
setpoint 40cm

IV. Conclusion and Future Enhancement

The first principle model of the variable area tank system has been determined using empirical
approach. A nonlinear model based control scheme have been designed and implemented on the
conical tank system. From the extensive experimental studies it can be inferred that the proposed
control scheme delivers satisfactory servo and regulatory responses than a gain scheduled IMC
based PI controller. The controlling of nonlinear process is a challenging task.

The Model Based controller is designed in such a way that the system is robust and physically
realizable. But due to the presence of dead time, the performance of the system may be affected.
Using advanced control schemes such as Model Predictive control better performance and
robustness can be obtained.


1) “Design of Deadbeat Algorithm for a Nonlinear Conical tank system

Marshiana. Da* , Thirusakthi murugan” 3rd International Conference on Recent Trends in
Computing 2015 (ICRTC-2015) Science Direct.

2) “A new autotuning algorithm for PID controllers using dead-beat format” R.

Bandopadhyay , Dr. Patranabis, ISA Transactions 40(2001) 255-266.

3) “Experimental validation of a Nonlinear Model based Control scheme on the

Variable area tank process” S. Kapil Arasu, Atanu Panda, J. Prakash IFAC-Papers OnLine
49-1 (2016) 030–034, Science Direct.

4) “Model based Controller Design for Conical Tank System” D.Angeline , K.Vivetha ,
International Journal of Computer Applications (0975 – 8887) Volume 85 – No 12,
January 2014

5) “Deadbeat Response Design” – NPTEL

6) “Digital Control” Kannan M. Moudgalya Indian Institute of Technology, Bombay