Thesis Diseño de modelo estadistico para estimaciones de peso estandar para los equipajes de mano correspondientes a los vuelos de la linea aérea LAN.
Loading...
Date
2004
Authors
Journal Title
Journal ISSN
Volume Title
Program
Ingeniería en Aviación Comercial
Departament
Campus
Campus Santiago Vitacura
Abstract
Lan, compañía aérea comercial reconocida mundialmente y consolidada como una de las mejores líneas aéreas de Sudamérica, se ha destacado en su larga trayectoria por optimizar sus diferentes actividades, logrando entregar un servicio eficiente y de calidad a sus clientes. El compromiso que ha tenido la compañía a lo largo de su historia, ha sido responder con excelencia a la gran cantidad de clientes que depositan su confianza tanto en el área de pasajero como en el área de carga. Para esto es fundamental generar y entregar seguridad, la cual se debe ver reflejada en el servicio que ofrece la compañía como también en las operaciones que ésta tenga, garantizando la seguridad tanto de sus trabajadores como de sus clientes.
La seguridad en el vuelo, en el despegue y aterrizaje es la que genera la necesidad de evaluar los distintos pesos que pueden afectar estas operaciones. Dentro de estos pesos existe una incógnita del peso de equipaje de mano que se lleva a bordo del avión, el cual es importante considerar en el proceso de estiba del mismo. Se debe tener un cuidado especial, por motivo que una mala estimación del peso estándar del equipaje de mano puede implicar un mayor peso de carga facturada, y, por tanto, exceder el Payload permitido por el fabricante para mantener los estándares de seguridad.
Para ayudar a la seguridad aérea, Lan actualmente determinó controlar el equipaje de mano de los pasajeros, el peso de los tripulantes y de equipaje de mano de los tripulantes. En el desarrollo de este trabajo se especificará como se procede a tomar las muestras en diversos vuelos, para el equipaje de mano de los pasajeros, para posteriormente evaluar la forma en que se distribuyen.
Para determinar como se distribuyen los pesos de equipaje de mano se analizará el comportamiento de estos como una distribución Normal, la que es utilizada hoy en día por la FAA y que LAN también quiere adoptar a corto plazo.
El análisis de esta distribución se desarrollará gracias a la información obtenida de 50 vuelos, de índole tanto nacional como internacional. Toda esta información se analizará mediante gráficos, que detallarán si efectivamente los pesos se comportan como una distribución Normal.
Al analizar los comportamientos gráficos, se concluye que la distribución empleada por la FAA, no es apropiada para la línea aérea LAN, pero al contemplar la forma de distribución de los datos se determinó que su comportamiento corresponde a la distribución de Weibull, la cual es usada hoy en día para modelos de tiempo y fallas.
La función de densidad de Weibull contiene dos parámetros, α y β. El parámetro de escala β, refleja el tamaño de las unidades en que se mide la variable aleatoria γ. El parámetro α, es el parámetro de forma, el cual si es variado permite generar un conjunto con una amplia variedad de curvas.
Una vez establecidos los datos, para ver el comportamiento de estos como una distribución de Weibull, se efectúa la comparación de esta función con otra que determina el porcentaje de área por rango, para luego determinar el error de aproximación que tiene la distribución de Weibull en relación a la función de porcentaje de área por rango. Este mismo procedimiento se ejecuta para una distribución Normal, determinando cual tiene menor error de aproximación.
Al realizar la operación anterior se constata que el 100% de las muestras tomadas se comportan como la distribución de Weibull. Para garantizar que la función de Weibull es apropiada, se utilizó el supuesto de Kolmogorov, el cual es un test de hipótesis que determina la bondad y ajuste de las curvas.
Este estudio realizado, será un gran avance para la seguridad de los trabajadores, como para los pasajeros, lo que es fundamental para poder integrarse en el mercado como una empresa consolidada a nivel internacional en términos de confiabilidad.
The analysis of this distribution will be developed thanks to the information obtained from 50 flights, considering both domestic and foreign. All this information will be studied with graphs, which will detail if effectively the weights behave as a Normal Distribution. When analysing the behaviour of these graphs, it is concluded that FAA’s distribution is not appropriate for the airline LAN, but, when contemplating the structure of data’s distribution, it was determined that it belongs to a ‘Weibull Distribution’, which now days is used for time & fault models. Weibull’s density distribution contains two parameters, α y β. The scale parameter β shows the size of the units in which the random variable γ is measured. α is the shape parameter, which if variated, allows to generate a set with a wide variety of curves. Once data is established, in order to see it as a Weibull distribution, a comparison with other kind of distribution that determines the percentage of area per range is made, so the determine approximation error that Weibull distribution has, in relation with de area per range percentage function. The same procedure is applied for a Normal Distribution, establishing which one has fewer approximation error. After executing the previously described operation, it can be seen that 100% of the simples taken behave as a Weibull Distribution. To guarantee that Weibull function is appropriate, Kolmogorov supposal is used, which is a hypothesis test that determines curves kindness and adjustment. This study will be a great advance for workers and passengers security, what it is fundamental to be able to integrate itself in the market as a company consolidated at international level in trustworthiness terms. LAN, an air commercial company worldwide recognized and consolidated as one of the best airlines in South America, has outstanding on its long trajectory for optimizing its different activities and being able to offer an efficient and quality service to its customers. The commitment this company has had throughout its history has been to answer with excellence to all customers who rely and trust in both passenger and cargo areas. For this, its fundamental to generate and create safety, which must be reflected on the service provided and in operations as well, guaranteeing safety of its workers and its clients. Safety during flight, take-off and landing generates the need to evaluate de different weights that may affect these operations. Inside these weights, there is certain unknown that is the hand baggage taken inside the airplane. Special care must be taken, because a bad estimation of the standard hand baggage weight may imply a mayor payload, hence, exceed the maximum payload allowed by the manufacturer, to maintain the security standards. To help air safety, LAN determined to control weight of passengers and crew’s hand baggage. In the development of this thesis, the process of how to choose different samples on diverse flights for passenger’s hand baggage will be specified, and then evaluate the way that they are distributed. To determine the way that this hand baggage weights are distributed, they will be analysed as a Normal Distribution, which does FAA use now days, and that LAN also wants to adopt, in a short term.
The analysis of this distribution will be developed thanks to the information obtained from 50 flights, considering both domestic and foreign. All this information will be studied with graphs, which will detail if effectively the weights behave as a Normal Distribution. When analysing the behaviour of these graphs, it is concluded that FAA’s distribution is not appropriate for the airline LAN, but, when contemplating the structure of data’s distribution, it was determined that it belongs to a ‘Weibull Distribution’, which now days is used for time & fault models. Weibull’s density distribution contains two parameters, α y β. The scale parameter β shows the size of the units in which the random variable γ is measured. α is the shape parameter, which if variated, allows to generate a set with a wide variety of curves. Once data is established, in order to see it as a Weibull distribution, a comparison with other kind of distribution that determines the percentage of area per range is made, so the determine approximation error that Weibull distribution has, in relation with de area per range percentage function. The same procedure is applied for a Normal Distribution, establishing which one has fewer approximation error. After executing the previously described operation, it can be seen that 100% of the simples taken behave as a Weibull Distribution. To guarantee that Weibull function is appropriate, Kolmogorov supposal is used, which is a hypothesis test that determines curves kindness and adjustment. This study will be a great advance for workers and passengers security, what it is fundamental to be able to integrate itself in the market as a company consolidated at international level in trustworthiness terms. LAN, an air commercial company worldwide recognized and consolidated as one of the best airlines in South America, has outstanding on its long trajectory for optimizing its different activities and being able to offer an efficient and quality service to its customers. The commitment this company has had throughout its history has been to answer with excellence to all customers who rely and trust in both passenger and cargo areas. For this, its fundamental to generate and create safety, which must be reflected on the service provided and in operations as well, guaranteeing safety of its workers and its clients. Safety during flight, take-off and landing generates the need to evaluate de different weights that may affect these operations. Inside these weights, there is certain unknown that is the hand baggage taken inside the airplane. Special care must be taken, because a bad estimation of the standard hand baggage weight may imply a mayor payload, hence, exceed the maximum payload allowed by the manufacturer, to maintain the security standards. To help air safety, LAN determined to control weight of passengers and crew’s hand baggage. In the development of this thesis, the process of how to choose different samples on diverse flights for passenger’s hand baggage will be specified, and then evaluate the way that they are distributed. To determine the way that this hand baggage weights are distributed, they will be analysed as a Normal Distribution, which does FAA use now days, and that LAN also wants to adopt, in a short term.
Description
Keywords
Aerolíneas, Seguridad aérea, Lan Airlines, Modelos estadísticos
