Fuzzy Logic for pH Neutralization Process

pH neutralization process is a process that is widely studied due to its highly nonlinear
process reaction. Its nonlinearity behavior is caused by static nonlinearity between pH
and concentration. This nonlinearity depends on the substances in the solution and on
their concentrations. In this project, the nonlinearity of the process was investigated.
Later, the mathematical model of the process was developed based on McAvoy et al
[I]. In addition to the mathematical model, an empirical model was also obtained
from Analytical & Chemical Pilot Plant located in the Process Control &
Instrumentation Laboratory (23-00-06). Both models were then used to develop the
Fuzzy Logic Controller (FLC) by using Advanced-Neuro Fuzzy Inference System
(ANFIS) and also gain-scheduling method. In ANFIS implementation for empirical
model, the FLC output was identical to the output from PID. Therefore it is concluded
that FLC could be used to replace PID for empirical model. In ANFIS implementation
for mathematical model, the FLC also could be implemented for mathematical model
since the controlled variable successfully follows all the set point changes. For gainscheduling
method, the FLC was tested on servo and regulator problems. The servo
test was performed by using a random number generator to generate random pH set
points between 3 and 11 and the simulation is performed for 100 seconds. The result
for the servo test was similar with the result from the ANFIS implementation for
mathematical model. For regulator test, the disturbance was the ±20% variation in
acid flow. The result for the regulator shows, the controller manages to eliminate the
disturbance effect in the process variable. In overall, the project successfully shows
that FLC could be a good alternative to PID controller.