Control System Simulator for MIMO Tank Level using Self-tuning PID-Fuzzy Adaptive Controller

Sammanfattning: According to Oxford dictionary, the word intelligence is derived fromintellect, which is the ability of knowing, reasoning and understanding.Intelligent behaviour is thus the ability to reason, plan and learn, which in turnrequires access to knowledge.Intelligent control techniques that match characteristics of biological systemsproposition introductions for creating control products with new abilities.Since fuzzy logic was first presented by Prof. Lotfi A. Zadeh, the number offuzzy logic control applications has increased dramatically. For example, in aclassical proportional, integral, and differential (PID) controller, what ismodeled is the system or process being controlled, while in a fuzzy logiccontroller (FLC), the focus is the human operator’s behavior [9].This project includes two models with design and simulation of a tank levelprocess and experimental estimation of closed loop control system. Fluid levelhas an important part in the process industries, specifically in chemical andnuclear plants, and the controlling of fluid level at desired set point value is abig task in control strategy.The first model is simulating a single-input single output (SISO) tank levelaccording to the physical dimensions of the tank system. The second model ismore complex and simulating a multi-input multi output variables (MIMO)for coupled tank system. The MIMO has a nonlinear characteristic that islinearized by using Jacobian approximation. The models and control systemshave been developed in LabVIEW.The PID-Fuzzy controllers are generally more advanced than the conventionalones, mostly for higher-order, time-delayed, nonlinear systems and for thosesystems that have only uncertain mathematical models which the classic PIDcontrollers are difficult to handle.For lower-order linear systems, it will be seen that both conventional andPID-Fuzzy controllers work equally well.The results from this work confirms that, self-tuning PID-Fuzzy adaptivecontroller has a very high performance in comparison with fixed PIDparameters. The simulation in platform of LabVIEW show that PID-Fuzzycontroller is more robust than a PID controller, has a good response speedwith a high stability, no overshoot, no transient between interacted coupledtanks, and the controller has tested many times without observation, e.g., nodisturbance when the level suddenly changes between the tanks.