Generalized Minimum Variance Control for Water Distribution System

This paper presents the project of an adaptive control system to modeling and controlling, in real time, a water distribution system. The controllers act in determining the turning speed of the motor/pump set and in the opening angle of control valves. An experimental workbench was set aiming at modeling and identifying parametrically a hydraulic network. The control system is a Generalized Minimum Variance self-tuning (GMV), whose parameters where real-time estimated by the recursive least squares (MQR) method according to the imposed inputs to the converter and the valves. The GMV aims at the pressure control in measuring points, which provided the reduction of energy and water in the workbench. Experimental results show the performance of adaptive control system and its ability to control the water distribution process.