Modelling and Control of Open-Flow Canal Systems using LPV Techniques

Autor: BOLE MONTE, YOLANDA

Fecha de lectura: 05-11-2010

Resumen de la tesis

The lack of water will be one of the most important problems of the Planet in the next decades. In 2025, the world’s population will increase by 1.8 billion people (depending on the sources), specifically in the regions less developed from the perspective of the hydraulic technology and, the automatic and control engineering. In parallel, the most part of the western countries have acquired some irresponsible habits in the water consumption, so that the supply of water, the farming and the treatment and distribution of water resources are in danger. Then, an efficient usage of water is necessary as well as water resources protection. Water demand is growing in the last years and about the 70-80% of water is used for irrigation in agriculture. Therefore, around 40% of the population in the world live in countries where is actually difficult or almost impossible to have access to fresh water to satisfy the nutritional, industrial and domestic needs.

Therefore, the most developed countries utilize financial resources to improve irrigation systems management, that is, in an efficient manner. Irrigation canals are systems that are used to carry water from a source (a river, a reservoir) to the final users (plot of land, pumping station). Often, these systems are formed generally by a main canal, a series of secondary distribution canals and a set of regulation elements (sensors, sluice gates). Traditionally, canals management was done purely by manual operation of regulation elements and other structures. This implied an actuation based only in the local information of the canal and highly dependent on the capacity, experience and mobility of the operation staff. These limitations, together with the complexity and extension of a canal network, explain the lower efficiency in the distribution related to its potential performance.

Although it is possible to improve the traditional canal management introducing automatic control. Canals are dynamic systems with distributed parameters with a vast spatial distribution, with a varying dynamics depending on the operating point, strong delays (that vary with the operating point) between the control action and the observed response, with non-linearities, coupling and interactions among pools, etc. Specifically, this PhD thesis aims to LPV modeling and LPV techniques control based on LMIs fulfilling the stability and performance specifications over any operating point.