Semi Physical Modeling and Control of Nonlinear System Using Soft Computing Technique

In Chemical industries level control of non linear interacting and non interacting systems is on huge demand due to multivariable process interactions. In this work a non linear liquid level process represented by a 40 liter four non interacting spherical tank was subjected to dynamic analysis. The data was indentified to be non linear and approximated to first order model with an error of less than 5percent.The level was measured using on– line Honeywell capacitance sensor. From the model parameters, PI and fuzzy tuned PI Controller s were designed using MATLAB. The closed loop performance was studied for both servo and regulator problems. Based on overshoot, rise time, settling time, and ISE, it is found that the Fuzzy tuned PI controller is better suited for this process.