Simulative and Modeling of a Proportional Integral Derivative Controller to Access Body Temperature

ABSTRACT:
Hypothermia has been recognized as a significant contributor to neonatal morbidity
and mortality for all newborn infants. Hypothermia occurs when the body temperature drops
below 36.50
C. The normal body temperature is between 36.50
C and 37.50C. However, several
control algorithms have been Developed in the literatures to control the temperature of now
infants placed in a room or an incubator. Conventional PID controller is one of the control
algorithms that have been employed but not very suitable in temperature control system due
to complexity of the actual temperature control system, varied parameters, large inertia and
large delay. Therefore, a neuro-fuzzy adaptive PID temperature controller using neuro-fuzzy
reasoning method to auto-tune PID parameters was developed.
The model was developed in MATLAB simulink environment. The actual
temperature of the room was measured by the sensor inside the room and compare with the
normal temperature of the neonates through the comparator to know the error. The error, as
well as a change in error of body temperature was passed to the Artificial Neuro-Fuzzy
Inference System (ANFIS) controller. The ANFIS incorporates the features of both fuzzy and
neural network. The neural network aspect of this work reduced the searching space and time
for achieving an optimal solution while the fuzzy logic aspect applied linguistic rules on this
input to produce three outputs which was defuzzified to have PID gains (i.e. proportional gain
Kp, integral gain Ki and derivative gain Kd). These PID gains were passed to the PID
controller and the output of the controller was used to control the temperature of the room
where the neonate is kept. The development of mathematical modeling of the neonate's room
was achieved using law of conservation of energy and then designed the Neuro-Fuzzy PID
controller.
In conclusion, this project will develop an improved temperature control algorithms
for controlling neonate temperature room using ANFIS-PID which is believed to have a
better efficiency of ANFIS for tuning the PID gains over Fuzzy-PID with shortest settling
time and minimum overshoot.