Ciencias Exactas y Ciencias de la Salud
Permanent URI for this collectionhttps://hdl.handle.net/11285/551039
Pertenecen a esta colección Tesis y Trabajos de grado de las Maestrías correspondientes a las Escuelas de Ingeniería y Ciencias así como a Medicina y Ciencias de la Salud.
Browse
Search Results
- Quality 4.0 methodology for manufacturing processes(Instituto Tecnológico y de Estudios Superiores de Monterrey, 2024-12-03) Macias Arregoyta, Daniela; Morales Menéndez, Rubén; emipsanchez; Vázquez Lepe, Elisa Virginia; School of Engineering and Sciences; Rectoría Tec de Monterrey; Escobar Díaz, Carlos A.In this work, the pathway for the implementation of Quality 4.0 is reviewed. Several articles are written to detail the evolution from the execution of Six sigma DMAIC methodology and how it can be adapted to the use of AI to create smart factories and smart processes. The main objective is to expand the current conformance rate of this methodology and find the defective items that can be overlooked in manufacturing processes. This research ranges from the selection of the data to train the available models, how can it be corrected and improved, the different processes to handle real-world data sets, the use of different ML algorithms for data analysis, the adaptation of this MLAs to quality standards in Quality 4.0 practices, to the curricular needs for Quality 4.0 for problem solving replacing Six Sigma practices. This thesis will focus only on the description of the decay of the Six Sigma DMAIC paradigm and its evolution to Quality 4.0, and the Process Monitoring for Quality methodology for rare event detection, which are the most noted journal papers in which I could collaborate.
- COVID-19 mortality prediction using deep neural networks(Instituto Tecnológico y de Estudios Superiores de Monterrey, 2022-06) García Zendejas, Arturo; MORALES MENENDEZ, RUBEN; 30452; REPOSITORIO NACIONAL CONACYT; Morales Menéndez, Rubén; emipsanchez; School of Engineering and Sciences; Campus MonterreyCOVID - 19 disease caused by the virus SARS-CoV2 appeared in Wuhan China in 2019, in March 11th 2020 it was declared a global pandemics, taking by March 2022 over 5,783,700 lives around the world. COVID-19 spreads in several different ways, the virus SARS-CoV2 which causes COVID-19 can spread from a mouth or nose of a person who is infected through liquid particles whenever they cough, sneeze, speak or breath. Initial symptoms and development of the illness are catalogued as mild, because of that it may be difficult to identify which persons will more probably develop severe disease. One great support that can be given to medical centers and healthcare workforce would be the ability to predict which patients will have a greater risk of death and would develop more quickly and severe illness, in order to make triage for treatment and decisions about resources distribution. Machine learning and specifically Deep Learning works by modelling hierarchical representations behind data, aiming to classify or predict patterns by stacking multiple layers of information. Some of its main applications are speech recognition, natural language processing, audio recognition, autonomous vehicles and even medicine. In medicine, it has been used to predict how a disease develops and affects patients. During this thesis it was done a research and comparison of state of the art articles and models that aim to predict the behavior and development of COVID-19 patients and the illness itself. Their different datasets, metrics, models and results have been studied and used as a base in order to create the proposed models of the thesis. This research project proposes the use of machine learning models to predict the mortality of COVID-19 patients by using as input attributes of the patients such as vital signs, biomarkers, comorbidities and diagnostics. This data was obtained for training and testing purposes from different medical centers, such as HM Hospitals, San Jose Hospital and CEM Hospital. The main Deep Learning model used during this thesis is a Deep Multi-layer Perceptron Neural Network which uses static attributes, and a Long-Short Term Memory Recurrent Neural Network using dynamic attributes. A mixed model combining the static and dynamic model was also created. It was also used metrics that support the reduction of false negative cases, the Maximum Probability of Correct Decision is the main metric to evaluate and optimize the model. The models have been evaluated and compared with another machine learning models such as Random Forest and eXtreme Gradient Boosting over the different datasets.
- LSTM Neural Networks for Remaining Useful Life Estimation of Turbofan Engines(Instituto Tecnológico y de Estudios Superiores de Monterrey, 2020-12-04) Montoya Herrera, Luisa Fernanda; MORALES MENENDEZ, RUBEN; 30452; Morales Menéndez, Rubén; tolmquevedo, emipsanchez; School of Engineering and Sciences; Campus MonterreyCondition-based Maintenance is a maintenance strategy that monitors the actual condition of a system to make predictive decisions whit respect to it. This type of maintenance includes detection, diagnosis, and prediction of system failures. It has become increasingly important because it generates the least losses, reducing total maintenance costs in a business by 5In general, the Remaining Useful Life estimation allows making failure predictions. The complexity of failure prediction in mechanical systems has led to a significant amount of literature. Different solutions have been proposed; however, this still a real problem.Remaining Useful Life estimation can be done from other approaches, for example, using physical models, knowledge-based models, or data-driven models. Extracting relevant features from raw data using physical or knowledge-based techniques alone, in most cases, is not enough due to the complexity of the characteristics present in the data. Literature shows that data-driven approaches are the most used for prediction. In recent years, Deep Learning models for different applications have been used, including failure detection, diagnosis, and prediction. The Deep Learning model’s advantage is that an indepth knowledge of the system is not required, and due to its robustness, complex learning results are satisfactory. For Remaining Useful Life estimation, Long Short Term Memory neural networks are a viable option since they can adequately handle the time series needed for failure predictions using Remaining Useful Life estimation. The three main stages for developing this method based on Long Short Term Memory neural networks were data pre-processing, model training, and model performance evaluation. The methodology uses two datasets of turbofan engines with different operational conditions and faults for its validation. The process evaluates signals obtained from sensors located along with a turbofan engine simulated through a Simulink-based program. This methodology presents a reasonably acceptable performance in terms of Root Mean Squared Error of 2.85 with a standard deviation of 0.39. It means that on average for the engines, the failure prediction will have an error of 3 cycles; and a Score function of 7.26 with a standard deviation of 1.76, which is an asymmetric algorithm where late predictions are more penalized than early predictions, increasing exponentially with the error. The proposed methodology has the advantage of being more straightforward than other methods found in the literature. Besides, the obtained values of the predictions are conservative.
- Human-vehicle interaction analysis(Instituto Tecnológico y de Estudios Superiores de Monterrey, 2020-06-05) Campos Ferreira, Andrés Eduardo; MORALES MENENDEZ, RUBEN; 30452; Morales Menéndez, Rubén; ilquio, emipsanchez; Vargas Martínez, Adriana; Ramírez Mendoza, Ricardo Ambrocio; School of Engineering and Sciences; Campus Monterrey; Lozoya Santos, Jorge de JesúsThe present work is the Thesis work of Vehicle assessment comparison from a smartphone reference with different approaches, to pursue the Master on Science on Manufacturing Systems. The automotive industry is continuously evolving by implementing top-edge technologies to improve comfort, safety, and driving experience to the users. In the context of Industry 4.0 and the Smart Cities paradigm, the concept of Intelligent Transportation System has become a research topic in the last few years. In the race for autonomous driving, researchers and industry have stressed the importance of monitoring drivers and passengers to determine the driving style, safety, and fuel efficiency, among other essential features. Despite all the work that has been done to monitor drivers, some approaches consist of vehicle-fixed devices or personalized devices that do not allow for the reproduction of experimentation to other vehicles. This instrumentation limits enormously the possibility to monitor any type of vehicle and collect information to develop intelligent algorithms that can predict driver and vehicle features such as driving behavior, energy consumption, fatigue, or vehicle’s element prognosis. Current researches focus on analyzing the interaction as a system from the vehicle’s point of view or driver’s point of view. Nevertheless, they have not been observed on both sides. To overcome these issues, an experimental setup is proposed on this work. The importance of this project is the easiness and replicability of the experimental setup; it is then validated by analyzing the logged data and the correlations between variables. Besides, state-of-the-art algorithms are compared to validate and select the best performance. This thesis integrates an experimental setup easy to use and implement with available commercial devices. Then, to validate the setup, a selection of algorithms based on a literature review were replicated and fed with the data logged from the experimental setup. A set of analyses of the resulting dataset is done to observe the interaction of vehicle and driver signals’ performance on how these signals are correlated. The first part of this work is devoted to the experimental setup definition and testing. Here the process was iteratively done by generating a procedure. Then, the next step consisted of exploring the logged data with a statistical tool to determine a possible correlation between signals and to reduce the dataset order but preserving most of the information. Later, state-of-the-art algorithms and data-driven identification models were identified and validated for specific key performances of vehicles and drivers. The vehicle’s key performances boarded on this thesis are the driving style, energy consumption, and emissions. Besides, the driver’s key performances are the heart condition, temperature, electrodermal activity, and heart rate. These features are highly studied when evaluating vehicle’s or driver’s state. The result of evaluating these performances with the selected algorithms shown that the driving style had 77% of correct classification, energy consumption, and emissions had around 16% and 11% of relative error, respectively. The results of this project show how vehicle and driver interact by analyzing its key performances. The Principal Component Analysis technique helped to find correlation among the raw data and also reduced the features from 57 to 27 without significant losses on the information. Besides, it demonstrated the correlation between vehicle’s and driver’s key performances by analyzing PCA plots and the covariance matrix.
- Control of semi-active suspensions for in-wheel electric vehicles(Instituto Tecnológico y de Estudios Superiores de Monterrey, 2020-06) Anaya Martínez, Mauricio; MORALES MENENDEZ, RUBEN; 30452; Morales Menéndez, Rubén; emipsanchez; Vargas Martínez, Adriana; Ramirez Mendoza, Ricardo Ambrocio; Escuela de Ingeniería y Ciencias; Campus Monterrey; Lozoya Santos, Jorge de JesúsWith the electric vehicles highly adoption, there is a need for keeping improving automotive systems. This work is focused on exploring the use of semi-active suspension systems in in-wheel electric vehicles. For that, two different in-wheel concepts are considered. When a brush-less DC (BLDC) motor and when a switched reluctance motor (SRM) are employed. In the SRM, an unbalanced vertical force is taken into account for the vertical dynamics model. The vertical dynamics tests are performed making use of models from one-quarter of vehicle (QoV) andf ull vehicle. Four different semi-active controllers, as well as three current levels, are evaluated and compared in time and frequency domain when employed in the in-wheel and internal combustion engine (ICE) vehicles. The suspension objectives improvement is estimated by making use of some performance indexes. Where the obtained results are compared against the ones given by the 1.25A and F-class baseline suspensions. The results showed that when compared against the F-class baseline suspension, none of the controllers is giving human and ride comfort improvements for the in-wheel electric vehicles. While, in comparison with the 1.25A baseline, the FEB controller is providing the best increase (25%−50%). By the side of the road holding and handling, the M1S guarantees the road holding and handling improvement (10%-25%) for the BLDC. While for the SRM, the FEB controller improves them when compared against the F-class baseline suspension. When taking as reference the 1.25A baseline suspension, the road holding and handling are enhanced by the 1.25A baseline and GH, respectively, for the BLDC. While in the SRM, the FEB controller is giving the best improvement(10%−55%). In most cases, high and low current values guarantee the comfort and road holding improvements, respectively.
- Hilbert-Huang transform based methodology for bearing fault detection(Instituto Tecnológico y de Estudios Superiores de Monterrey, 2018-05-16) Campos García, Rubén; Vallejo Guevara, Antonio Jr.; Morales Menéndez, Rubén; Ibarra Zárate, David IsaacRotating machinery is of great importance for manufacturing industry, and therefore huge investments for their acquisition are made every year. Machine preservation plays an important role in the exploitation of this resource. Rotating machines are more susceptible to certain types of faults, investigations report that at least 42 % of the root causes of failure in rotating machinery are related with bearings. To detect the bearing condition many techniques have been developed. One of the most reliable is vibration analysis. The Hilbert-Huang transform (HHT) has been used for vibration analysis and has gained attention in recent years, a topic of controversy in this method is the selection of the Intrinsic Mode Functions (IMFs) with fault information. Statistical parameters can be used to describe the characteristics of vibration signals, this attribute can be exploited to select the IMFs. There are many time domain features used for signal analysis. In this research, a study of 17 statistical parameters was made to determine which one is the best to represent IMFs with fault information. As a result of this analysis a new methodology based on HHT is proposed. This methodology deals with the IMF selection with the use of KR (Kurtosis x RMS) to detect the IMFs with fault information, and can be used to detect incipient bearing faults. The proposed methodology was validated with 18 signals from the Case Western Reserve University (CWRU), Tian-Yau Wu, and the society for Machinery Failure Prevention Technology (MFPT Society) databases. For the 18 analyzed signals, only one IMF was wrongly selected. The cause of this error was the end defect produced in the EMD, this caused the KR amplitude to increase even tough the IMF did not have fault information. The results on the Envelope spectrum from 14 signals were clear with fault components with large amplitude. For the remaining four signals the results on the Envelope spectrum was noisy, but the fault fault components were distinguishable.Rotating machinery is of great importance for manufacturing industry, and therefore huge investments for their acquisition are made every year. Machine preservation plays an important role in the exploitation of this resource. Rotating machines are more susceptible to certain types of faults, investigations report that at least 42 % of the root causes of failure in rotating machinery are related with bearings. To detect the bearing condition many techniques have been developed. One of the most reliable is vibration analysis. The Hilbert-Huang transform (HHT) has been used for vibration analysis and has gained attention in recent years, a topic of controversy in this method is the selection of the Intrinsic Mode Functions (IMFs) with fault information. Statistical parameters can be used to describe the characteristics of vibration signals, this attribute can be exploited to select the IMFs. There are many time domain features used for signal analysis. In this research, a study of 17 statistical parameters was made to determine which one is the best to represent IMFs with fault information. As a result of this analysis a new methodology based on HHT is proposed. This methodology deals with the IMF selection with the use of KR (Kurtosis x RMS) to detect the IMFs with fault information, and can be used to detect incipient bearing faults. The proposed methodology was validated with 18 signals from the Case Western Reserve University (CWRU), Tian-Yau Wu, and the society for Machinery Failure Prevention Technology (MFPT Society) databases. For the 18 analyzed signals, only one IMF was wrongly selected. The cause of this error was the end defect produced in the EMD, this caused the KR amplitude to increase even tough the IMF did not have fault information. The results on the Envelope spectrum from 14 signals were clear with fault components with large amplitude. For the remaining four signals the results on the Envelope spectrum was noisy, but the fault fault components were distinguishable.
- Magneto-rheological damper modeling using LPV systems(Instituto Tecnológico y de Estudios Superiores de Monterrey, 2011-01-01) Díaz Salas, Vicente Alberto; DIAZ SALAS, VICENTE ALBERTO; 297034; Morales Menéndez, Rubén; tolmquevedo; Tecnológico de Monterrey, Campus MonterreyThe present research is focused in the dynamical modeling of a Magneto-Rheological Damper as a semi-active actuator. This device shows a complex behavior including non-linearities and hysteresis, features that are to be emulated by a dynamical system, in order to express mathematically the conduct of this device under mechanical vibrations. The M R damper is part of a semi-active suspension system, and by the use of the information of the model designed in this thesis it might be possible to produce a control design for the quarter of car system, in order to regulate the vibrations received by the suspension, incrementing the comfort and safety of the passengers in the vehicle. In order to obtain experimental data from this semi-active device, a set of tests were performed, introducing displacement and current excitation patterns into the shock absorber and measuring the dynamical response of it. Using the experimental results, a set of state of the art models (Semi-Phenomenologial, Phenomenological, Black-Box) were learned to reproduce this data. This work proposes an LP V (Linear Parameter-Varying) system, as a model for the M R damper, which is capable of reproduce both the non-linear and hysteretic behavior of the damper.The LP V model proposed has the capacity to create a relationship between the main excitation variables, and the damper force, in a single structure. Due to this feature, it might later be added to a bigger strategy, such as for control or observation. The model was designed using an LP V polynomial system and an switching variable, which depends on the input velocity and current. Output results shows a higher accuracy from the LP V proposed model, in comparison with the state of the art models reviewed: it has ESR index average values below 0.04 while most of the studied models only achieve values below 0.1. The proposed model of this thesis provides a dynamical description of the Magneto-Rheological damper that generates a link between the main input variables implicated on the damper, and the output force. This feature might become an advantage to later provide an extended model of the quarter of car, and design a controller to regulate the vibrations applied to the suspension.
- Metodología de diseño y desarrollo de equipo didáctico basada en inteligencia competitiva y tecnológica(Instituto Tecnológico y de Estudios Superiores de Monterrey, 2009-12-01) Fuentes Rivera, Laura Yazmín; Morales Menéndez, Rubén; Rodríguez Salvador, Marisela; tolmquevedo; Tecnológico de Monterrey, Campus MonterreyEn este documento de tesis se persigue como objetivo la descripción de una metodología de diseño y desarrollo de equipo didáctico basada en inteligencia competitiva y tecnológica, la cual es validada mediante una aplicación en el sector educativo. Este desarrollo se conforma de 5 capítulos: el primer capítulo inicia con la descripción del estudio en la introducción, planteamiento y justificación del problema; en el segundo capítulo se habla de los fundamentos teóricos que ayudan a comprender los conceptos centrales en los que se basa la metodología propuesta, así como de las herramientas necesarias para su implementación. Próximo a este capítulo se encuentra el capítulo 3, donde se describe a detalle la propuesta de tesis que toma como base los fundamentos revisados en el capítulo previo, ya que los conceptos revisados se integran para dar pie a la metodología de diseño y desarrollo de equipo didáctico, basada en inteligencia competitiva y tecnológica. En el cuarto capítulo se presenta la validación de la propuesta de tesis, la cual por ser aplicada en la educación tiene la finalidad de brindar un aprendizaje activo a los usuarios que empleen equipo didáctico. Es decir, en este capítulo se presenta el caso de estudio sobre la estación didáctica de automatismos lógicos de control, un instrumento de enseñanza que se utiliza en el laboratorio de control lógico, para estudiantes de mecatrónica en el Instituto Tecnológico y de Estudios Superiores de Monterrey, Campus Monterrey. Así, la última parte del documento (capítulo 5), muestra las conclusiones sobre la metodología que se propone y sobre los resultados de su aplicación en la educación.
- Diagnóstico de Fallas Usando el Análisis de Componentes Principales y de Correspondencia-Edición Única(Instituto Tecnológico y de Estudios Superiores de Monterrey, 2009-05-01) Celina Rea Palacios; Morales Menéndez, Rubén; Verde Rodarte, Cristina; Ramírez Mendoza, Ricardo A.; Tecnológico de Monterrey, Campus Monterrey; Acevedo Mascarúa,JoaquínEn un equipo industrial como un intercambiador de calor, las fallas en sensores y actuadores degradan el desempeño del sistema y pueden causar pérdidas económicas importantes. Ya que un sistema industrial puede contener una gran cantidad de sensores y actuadores, es necesario apoyarse en sistemas computacionales para realizar esta tarea. Existen diferentes métodos para la detección y diagnóstico de fallas (FDD), los métodos de detección basados en datos históricos son un ejemplo de estas aplicaciones. En esta investigación, se implementaron dos métodos de este tipo para su comparación y análisis: Análisis Dinámico de Componentes Principales (DPCA) y Análisis de Correspondencias (CA) en un mismo equipo industrial. Ambos métodos requieren representar los proceso vía modelos estadísticos calculados a partir de datos históricos del proceso en condiciones normales de operación y obtener umbrales escalares que capturen la correcta operación del proceso. Esto se logra en dos subespacios: El espacio principal y el residual, seleccionando sus componentes al obtener las gráficas de variación acumulada de información. Los estadísticos T2 de Hotelling y el Q se utilizaron para detectar una falla. El diagnóstico se puede realizar con las tablas de contribución residual. Los dos métodos se probaron con fallas abruptas introducidas en los transmisores, adicionalmente en DPCA se probaron fallas en actuadores. Se analizó el efecto que tiene la cantidad de componentes necesarios en cada subespacio, sensibilidad de los umbrales, la influencia del tiempo de muestreo y la cantidad de datos incluidos en el paso de entrenamiento, además de la efectividad del diagnóstico realizado con las tablas de contribución residual. Adicionalmente, para DPCA se analizo el efecto en la detección de la cantidad de retrasos en el modelo estadístico. Como resultados principales, en esta aplicación se encontró los componentes retenidos son constantes aunque cambie la dinámica del sistema, siendo 5 componentes principales para DPCA y 3 componentes principales para CA. Con las curvas Receiver Operating Characteristics se encontró que el estadístico T2 fue robusto para ambos métodos, el estadístico Q sólo para DPCA, mientras que éste mismo para CA se sugiere no monitorearlo al presentar falsas alarmas. El sistema se puede muestrear hasta 5 segundos sin generar falsas alarmas y la cantidad de datos para generar el modelo estadístico fue de 1500 a 2000 datos para DPCA y CA respectivamente. DPCA tiene 100 % de acierto en los diagnósticos, mientras CA tiene 90.9 % en el transmisor de temperatura de salida y 46.66 % en el flujo de agua. Para DPCA, el incluir más retrasos permite aumentar la rapidez de diagnóstico (Probabilidad de detección=[0.98,0.993]) en ambos estadísticos, pero también aumentan las falsas alarmas para T 2 (Probabilidad de falsas alarmas=[0,0.09]). Para aplicaciones similares se recomienda usar DPCA y CA como sistema de detección y diagnóstico de fallas considerando las sugerencias descritas en este documento, las cuales brindarán la detección más pronta y la mínima cantidad de falsas alarmas.
- Detección y diagnóstico de fallas en sensores de un motor de corriente directa(Instituto Tecnológico y de Estudios Superiores de Monterrey, 2009-01-01) López González, Miguel Ángel; Morales Menéndez, Rubén; Acevedo Mascarúa, Joaquín; Guedea Elizalde, Federico; Aguilar Coutiño, Artemio; Programa de Graduados en Mecatrónica y Tecnologías de Información; División de Mecatrónica y Tecnologías de Información; Campus MonterreyEl área de detección y diagnóstico de fallas está ligada directamente al diseño de nuevos sistemas o procesos, debido a los altos índices de seguridad requeridos hoy en día. Detectar fallas a tiempo puede evitar daños graves a los procesos, maquinarias e inclusive a seres humanos. Se propone un sistema de Detección y Diagnóstico de Fallas (FDD por sus siglas en inglés, Fault Detection and Diagnosis) basado en observadores para fallas aditivas en sensores. Tres diferentes tipos de fallas (falla abrupta, ruido en el sensor y falla gradual) son introducidas; la falla abrupta causa que el sensor presente un sesgo en la lectura. La falla por ruido causa que la salida del sensor presente ruido grande. La falla gradual se presenta normalmente por el desgaste de componentes electrónicos o mecánicos debido al uso del sensor por largos periodos de tiempo; causando que la salida del sensor varia lentamente. Se propusieron dos modelos lineales para representar el comportamiento de la corriente y velocidad en el motor sin falla, estos modelos fueron utilizados para el diseño del observador que es el generador de residuos. Las fallas se detectan y aíslan con un banco de observadores, el diagnóstico se lleva acabo analizando el comportamiento de la media y desviación estándar de los residuos generados por los observadores. El desempeño del sistema FDD propuesto para detectar y diagnosticar las fallas se pone a prueba en sensores utilizados para monitorear un motor de corriente directa colocado en el módulo del volante de un sistema de Dirección por Cable (SBW por sus siglas en inglés, Steer by Wire). El sistema FDD utiliza solamente la señal de excitación que es el voltaje, y la salida de los sensores corriente y velocidad. Los resultados mostraron que el sistema FDD logró detectar todas las fallas en aproximadamente 1 segundo para el sensor de corriente bajo las condiciones de operación establecidas y para las fallas propuestas, el tiempo de diagnóstico fue de 2 segundos debido al procesamiento del cálculo de los índices estadísticos. Para el sensor de velocidad se detectaron 4 falsas alarmas causadas principalmente porque el sistema opera en lazo abierto y las perturbaciones afectan severamente a la velocidad a la que gira el motor de CD.
