Sliding-mode control for an artificial neural network analysis and design of treatment water system

Abstract EN

This paper presents data base and artificial neural network (ANN) for ultraviolet (UV) treatment water system.

This latter constitutes the main part of a pilot water disinfection unit, which is able to distribute the supplied energy towards the receivers..

The UV water disinfection pilot unit, object of the present study, is fed by a photovoltaic (PV) source.

The PV generator depends on the solar irradiation.

This fluctuation of climatic conditions affects the result of water disinfection, which depends on the predetermined UV dose.

For this reason we had resorts to prepare a data base from bacteriologist’s analyses able to fix the flow rate and UV irradiation necessary according to the supplied energy to have the best water quality.

The ANN will allow us to control the system in real time insuring optimal water disinfection by UV light.

The non linear model relates the current and the conductivity of the low pressure discharge lamp used for water treatment was developed.

The sliding mode control approach of a class of non linear systems subjected to control approach for the power UV lamp and power motor-driven pump is established.

Reliable results are given to show the interest of the sliding mode control of complex physical systems.