The use of phenomenological based semi-physical models as virtual sensors for density and viscosity of mineral slurries

Abstract: this thesis presents two estimation structures for on-line estimation of density and viscosity of mineral slurry flowing through a pipes and fittings assembly (PFA). The structures are developed from a phenomenological based semi-physical model (PBSM) as main component of a soft-sensor. The PBSM is developed based on the mass conservation and momentum transfer principles, in order to separate the density estimate from the viscosity, and momentum transfer, considering frictional energy losses. In this way, the model equations are obtained based on the variables directly affecting the density and viscosity of minerals slurries. The proposed model is obtained in state variables, considering the available measurements (pressure drop and volumetric flow) and the disturbances in the process. In this sense, from the theory for state observers, an observer of unknown inputs was used in the estimation structure to take into account the inherent disturbances in the process inlet. The density is estimated as a state of the model while the two options for estimating viscosity are proposed based on the availability for characterizing the rheology of the mineral slurry from laboratory tests. Finally, the performance of the proposed structures for the estimation of the density and viscosity properties of the mineral slurry is tested by simulation. The conducted tests use disturbances in the mixture inlet composition and include noise in the measured variables. The results exhibit a good estimation of both slurry properties using both proposed methods.

• Autores: Diaz Castañeda, Jenny Lorena
• Fecha de publicación: 2017
• Institución: Universidad Nacional de Colombia
• Repositorio: Universidad Nacional de Colombia
• Idioma: spa
• OAI Identifier: oai:repositorio.unal.edu.co:unal/59135
• Acceso en línea: https://repositorio.unal.edu.co/handle/unal/59135; http://bdigital.unal.edu.co/56393/
• Palabra clave: 66 Ingeniería química y Tecnologías relacionadas/ Chemical engineering; Soft-sensor; Multiphase fluids; Phenomenological based semi-physical models; Loss of energy; State estimator; Sensor virtual; Fluidos multifásicos; Modelos semifísicos de base fenomenológica; pérdidas de energía; estimadores de estado
• Rights: openAccess
• License: Atribución-NoComercial 4.0 Internacional