Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → États-Unis → 2017 → Artificial intelligence and time series based forecasting in water resources, decision modeling, and optimal selection using ranking techniques

New Mexico State University (2017)

Artificial intelligence and time series based forecasting in water resources, decision modeling, and optimal selection using ranking techniques

Zamani Sabzi, Hamed,

Titre : Artificial intelligence and time series based forecasting in water resources, decision modeling, and optimal selection using ranking techniques

Auteur : Zamani Sabzi, Hamed,

Université de soutenance : New Mexico State University

Grade : Doctor of Philosophy (PhD) Civil Engineering 2017

Notes partiels
Beyond any doubt, water is the best present that every living organism can be given when it is not available around. In arid and semi-arid regions especially, it is crucially important to manage water optimally, and water management depends on accurately quantifying and predicting fresh water resources. To facilitate optimal water management, this study investigates optimal stormwater management, water forecasting, decision modeling, and optimal decision selection for water in the Mesilla Valley of southern New Mexico. Through interrelated investigations, this study explores the optimal management of stochastic floods in the Mesilla Valley ; develops and evaluates multi-criteria decision-making (MCDM) processes in water resource projects under uncertain (i.e., fuzzy) environments ; evaluates water resource forecasting by artificial intelligence and time series based techniques ; and develops a case study on stream flow forecasting models for the Rio Grande. The stream flow forecasting models are incorporated into the integrated system of Elephant Butte and Caballo reservoirs, where minimizing the total evaporation loss from the reservoirs is critically important. In this regard, in chapter 2, the optimal management of potential stormwater in Mesilla Valley, southern New Mexico was investigated. About 76 potential watersheds were recognized, where stochastic floods on those watersheds, would be managed efficiently. As a role model to the other watersheds, a dynamic operating system was successfully developed and applied for the flood control pond in Diez Lagos (DL), simulating the pond as a control volume. The same approach in simulating the stochastic floods is applicable to the rest of the watersheds throughout the western Mesilla Valley. For general water resource problems, in Chapter 3, this study analytically investigated and statistically compared the performances of ten commonly used MCDM techniques, using theoretical, programming, and simulation approaches. The findings facilitate optimal decision processes under fuzzy environments. Additionally, the study investigates streamflow forecast models for Rio Grande inflow to Elephant Butte Reservoir, and the forecast models are used to minimize evaporative losses from Elephant Butte and Caballo reservoirs, while preserving reliable water releases to downstream agricultural lands. Since most of the water resources projects are multi-criteria with different levels of uncertainty (fuzziness), and MCDM techniques mostly have been utilized to rank the potential solutions for water resources problems, through theoretical, programming, and simulation work, we developed the extensions of each individual selected MCDM technique from ten commonly used MCDM techniques in water resources, analytically investigated and statistically compared the performances of those techniques under fuzzy environment. In addition, to facilitate the selection of an appropriate MCDM method under a fuzzy environment, this chapter investigates and statistically compares the performances, similarities and dissimilarities of ten commonly used MCDM techniques. As such, the study developed in chapter 3 clarifies the application of optimal decision-making processes under fuzzy environments as well as detailed applications of decision-making techniques and their processes in water resources projects.

Présentation (NMSU Library)

Page publiée le 14 octobre 2018