Publication: IMPLEMENTACIÓN DE UNA ESTRATEGIA DE CONTROL
Date
2019
Authors
GUERRERO BARRÍA, VÍCTOR ANDRÉS
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Hoy en día, la utilización de molinos SAG representa un estándar para el proceso
minero de gran escala, en que los cicloconversores multipulso son parte medular de
estos accionamientos sin engranajes o “gearless” de alta potencia.
La potencia consumida de los equipos empleados en molienda se ha visto aumentada
debido a la economía de escala presente en plantas concentradoras de alta productividad,
por lo que se observa una tendencia evidente al aumento en el volumen de los
molinos utilizados y por ende a un aumento en la potencia que manejan los accionamientos
presentes en este tipo de aplicaciones.
El aumento en la potencia de los equipos instalados ha significado nuevos retos relacionados
a mejorar el desempeño eléctrico y mecánico de los accionamientos en
cuestión. Por ende, el desarrollo teórico de nuevas estrategias de control para drives
de alta potencia resulta imperante.
El desarrollo de este trabajo de investigación plantea mejorar el comportamiento eléctrico
de un cicloconversor de 12 pulsos de alta potencia, empleando una estrategia de
control predictivo. Las técnicas clásicas de control utilizadas a la fecha, la implementación de una estrategia de control predictivo para este tipo de convertidores y las
simulaciones que respaldan una comparativa acertada entre las distintas técnicas de
control, son tópicos discutidos detalladamente durante el desarrollo del documento
presentado.
El desarrollo de este trabajo se orientó al mejoramiento de la confiabilidad y desempeño de accionamientos gearless GMD de alta potencia empleados en molienda SAG
y de Bolas de alta capacidad, contribuyendo al estado del arte con resultados en las
líneas:
Detección no-invasiva de fallas en CCV y la operación tolerante a fallas
Análisis de fallas de accionamientos de alta potencia
Control predictivo de accionamientos con CCV de alta potencia.
Modelado y simulación para el desarrollo de un emulador de tiempo real de un
CCV de alta potencia.
Estos resultados se han establecido en publicaciones y en una patente de invención.
In modern high-power industrial processes, multipulse cycloconverters are fundamental to feed synchronous machines. Given the importance of these critical processes, especially for the grinding stage in the mining industry (semi-autogenous and ball mills), the reliability of the high power drive is critical. Any development relating to new control techniques applied to well-known naturally commutated converters are quite valuable, because they may result in an improvement in the electric and mechanical behavior of these high-power gearless applications, e.g. a reduction in harmonics distortion, energy consumption, oscillatory torques, abnormal operations, etc. Many predictive control schemes have been proposed for force-commutated converters, e.g. matrix converters, where the predictive strategy has shown promising results. This thesis work is focused on presenting how a predictive control strategy can be applied for a 12-pulse cycloconverter, despite the natural commutation characteristic of the converter in question. The proposed predictive control algorithm, predictive model of the PMSM and a comparison between the classic and predictive control are detailed. This research work was orientated on improving the confiability and performance of the high power grinding mill drives (or GMD) in semi-autogenous and ball grinding. This thesis contributes to the state-of-the-art in the following research topics: Non-invasive failure detection system and fault tolerance capability on high power Cycloconverters. Fault analysis in high power converters. Predictive control for high power drives. Modelation and software programming of a high power CCV in a real-time application. These results have been published and they led to an invention patent.
In modern high-power industrial processes, multipulse cycloconverters are fundamental to feed synchronous machines. Given the importance of these critical processes, especially for the grinding stage in the mining industry (semi-autogenous and ball mills), the reliability of the high power drive is critical. Any development relating to new control techniques applied to well-known naturally commutated converters are quite valuable, because they may result in an improvement in the electric and mechanical behavior of these high-power gearless applications, e.g. a reduction in harmonics distortion, energy consumption, oscillatory torques, abnormal operations, etc. Many predictive control schemes have been proposed for force-commutated converters, e.g. matrix converters, where the predictive strategy has shown promising results. This thesis work is focused on presenting how a predictive control strategy can be applied for a 12-pulse cycloconverter, despite the natural commutation characteristic of the converter in question. The proposed predictive control algorithm, predictive model of the PMSM and a comparison between the classic and predictive control are detailed. This research work was orientated on improving the confiability and performance of the high power grinding mill drives (or GMD) in semi-autogenous and ball grinding. This thesis contributes to the state-of-the-art in the following research topics: Non-invasive failure detection system and fault tolerance capability on high power Cycloconverters. Fault analysis in high power converters. Predictive control for high power drives. Modelation and software programming of a high power CCV in a real-time application. These results have been published and they led to an invention patent.
Description
Keywords
CICLOCONVERSOR , DRIVE GEARLESS , MOLINO SAG