Feedback Control Strategy for a Dye Injection Process

MASTER OF ENGINEERING
(INDUSTRIAL AUTOMATION)-MIA
ME503 INDUSTRIAL PROCESS CONTROL SYSTEMS
ASSIGNMENT 2
V1.0
STUDENT NAME
STUDENT ID NUMBER
PROFESSOR (TUTOR)
DATE OF SUBMISSION

ASSIGNMENT 2- PROJECT MIDTERM
FEEDBACK CONTROL STRATEGY FOR A DYE INJECTION PROCESS.
INTRODUCTION
Dye injection systems are used in both large and small scale applications. The system is
used to dye clothing based on set instructions. The amount of dye input in the pipe with flowing
water determines the concentration of a color on the cloth. The mixer mixes the dye and the
water together to ensure proper consistency of the two inputs before releasing the liquid into the
pipe. Such a system requires a controller to ensure that the desired level of concentration is
obtained at the output [1]. The system has inputs and output as described in the table below,
Parameter Description
Input variables (a) manipulated variables
- location of the optical sensor (L)
- regulating valve (dye input rate): adjust valve position
(b) Disturbance variables
- mixer speed and noise variables
Output variables (a) measured variables
- distance of the optical sensor from the mixer
- dye input rate (milliliters per second)
- valve position of the regulating valve
(b) unmeasured variables
- water/dye concentration (Q)
Control structure

An open loop system can easily become unstable and it is important to use a feedback
loop which tests for errors. The feedback loop takes the input and compares it to the reference
input. The difference between the two process variables is the error of the system. Some of the
key causes of errors in such systems are disturbances from the external and internal environment
of the system. There are a number of control structure which can be implemented in such a
situation such as the proportional, proportional-integral, and the proportional-integral-derivative
controllers. These controllers can be implemented for system in the first order, second order, and
other higher orders. The proportional controller is used in the first order systems. It manipulates
a proportional gain constant which is used to alter the value of the output. The gain parameter is
multiplied with the system first order system to give a new yield to the system. The gain
parameter minimizes the steady state error but it does not eliminate the errors. The gain
parameter is denoted as, Kp. Adjusting the value increasingly minimizes the steady state error.
The value may be increased to a given value which is consider the optimum value which results
in a reduced amplitude and phase margin. Exceeding the value may cause the output to oscillate
during dead time or lag. The proportional-integral controller, on the other hand, eliminates the
steady state error completely. The control parameters are denoted as Kp and Ki. The controller is
implemented in areas where speed is not a performance factor. Unfortunately, the controller is
not able to predict future errors in the system and it is not able to reduce the rise time nor
eliminate the oscillations that may result when the value of the proportional gain parameter is
very high. The PID controller solves the issues that other controllers are unable to solve. It
guarantees optimum control dynamics to obtain zero steady state error, faster response which
implies a shorter rise time, no oscillations, no overshoots, and higher stability. The controller can
be implemented in systems of higher orders unlike the proportional controller which is limited to
first order systems only [2].
PROBLEM STATEMENT
The dye injection system has the dye injection and water as the inputs. The two meet at
the mixer and the mixer mixes the components to form a concentration which flows forward.
There is an optical sensor along the film which monitors the dye/water concentration. A black
dye is used to design the shirts which blends with the water and is mixed at the mixer. There is
need to have controllers that monitor the amount of dye that is allowed to flow through the
regulating valve. Another controller is needed to determine if the mixing process is correct and if