Improved Liquid Level Control Design Using Mamdani Fuzzy Inference System.

Abstract

In industrial applications such as those in petrochemicals, nuclear, etc. liquid level control is very important: hence we have elaborated a straightforward case of control scenario, that can further be applied in multiple of these areas by simply varying the design to suit each case. We have given preference to the application of fuzzy logic rather than other control methods – hence describing a lot of the background information about it. Then, we move forward to apply the Mamdani type Fuzzy Inference System in demonstrating how the logic works in controlling the liquid level in a single tank system; first using 3-rules, then, 5-rules. We have designed and simulated in MATLAB, the 3-rules and the 5-rules systems. Unlike previous works on these cases, our results show better choices in that, key performance parameters viz rise time, settling time and overshoot, are better than previous results we have seen in literatures. We achieve these better results by varying some parameters, initial conditions and improved Simulink designs; to explore the viability of our chosen method of control. This work also further proves that the FLC is a better controlling method in these areas of its application than other control methods.