Fuzzy Control of a Hydrodynamic Process

Hydrodynamic processes, part of many industrial and urban infrastructure systems, usually lack accurate models, hence introducing ambiguity in a controller design procedure. Operators and control engineers provide significant knowledge about successful control of hydrodynamic systems. Fuzzy controller, suitable to exploit and incorporate available knowledge, is designed for control of a benchmark coupled tank system. Successful fuzzy controller must cope with slow dynamics of the process, dead zone of an actuator, and unipolar control signal, when employed in reference tracking tasks. The designed fuzzy controller implements fuzzy inference system with 15 rules, two inputs and one output, as well as three adjustable gain parameters. The proposed controller is successful in reference tracking tasks and disturbance suppression. Experiments, organised as reference tracking tasks and implemented as computer simulations, support the analysis.