Examinando por Materia "Learning algorithm for multivariable data analysis"

Mostrando 1 - 2 de 2
  • No hay miniatura disponible
    Ítem
    Advanced fuzzy-logic-based context-driven control for HVAC management systems in buildings
    (Institute of Electrical and Electronics Engineers Inc., 2020-01-01) Morales Escobar L.; Aguilar J.; Garces-Jimenez A.; Gutierrez De Mesa J.A.; Gomez-Pulido J.M.; Morales Escobar L.; Aguilar J.; Garces-Jimenez A.; Gutierrez De Mesa J.A.; Gomez-Pulido J.M.; Universidad EAFIT. Departamento de Ingeniería de Sistemas; I+D+I en Tecnologías de la Información y las Comunicaciones
    Control in HVAC (heating, ventilation and air-conditioning) systems of buildings is not trivial, and its design is considered challenging due to the complexity in the analysis of the dynamics of its nonlinear characteristics for the identification of its mathematical model. HVAC systems are complex since they consist of several elements, such as heat pumps, chillers, valves, heating/cooling coils, boilers, air-handling units, fans, liquid/air distribution systems, and thermal storage systems. This article proposes the application of LAMDA (learning algorithm for multivariable data analysis) for advanced control in HVAC systems for buildings. LAMDA addresses the control problem using a fuzzy classification approach without requiring a mathematical model of the plant/system. The method determines the degree of adequacy of a system for every class and subsequently determines its similarity degree, and it is used to identify the functional state or class of the system. Then, based on a novel inference method that has been added to LAMDA, a control action is computed that brings the system to a zero-error state. The LAMDA controller performance is analyzed via evaluation on a regulation problem of an HVAC system of a building, and it is compared with other similar approaches. According to the results, our method performs impressively in these systems, thereby leading to a trustable model for the implementation of improved building management systems. The LAMDA control performs very well for disturbances by proposing control actions that are not abrupt, and it outperforms the compared approaches. © 2013 IEEE.
  • No hay miniatura disponible
    Ítem
    LAMDA-HAD, an Extension to the LAMDA Classifier in the Context of Supervised Learning
    (World Scientific Publishing Co, 2020-01-01) Morales L.; Aguilar J.; Chávez D.; Isaza C.; Morales L.; Aguilar J.; Chávez D.; Isaza C.; Universidad EAFIT. Departamento de Ingeniería de Sistemas; I+D+I en Tecnologías de la Información y las Comunicaciones
    This paper proposes a new approach to improve the performance of Learning Algorithm for Multivariable Data Analysis (LAMDA). This algorithm can be used for supervised and unsupervised learning, based on the calculation of the Global Adequacy Degree (GAD) of one individual to a class, through the contributions of all its descriptors. LAMDA has the capability of creating new classes after the training stage. If an individual does not have enough similarity to the preexisting classes, it is evaluated with respect to a threshold called the Non-Informative Class (NIC), this being the novelty of the algorithm. However, LAMDA has problems making good classifications, either because the NIC is constant for all classes, or because the GAD calculation is unreliable. In this work, its efficiency is improved by two strategies, the first one, by the calculation of adaptable NICs for each class, which prevents that correctly classified individuals create new classes; and the second one, by computing the Higher Adequacy Degree (HAD), which grants more robustness to the algorithm. LAMDA-HAD is validated by applying it in different benchmarks and comparing it with LAMDA and other classifiers, through a statistical analysis to determinate the cases in which our algorithm presents a better performance. © 2019 World Scientific Publishing Company.