بعض الطرائق المعلمية واللامعلمية لتقدير دالة المعولية مع تطبيق عملي == Some of The Parametric Methods And Nonparametric To Estimate The Reliability Function With The Practical Application
Author name:
بشير فيصل محمد حبيب السعدي
Supervisor name:
انتصار عريبي فدعم الدوري
General topic:
Administration and Economics
Specific topic:
Operations Research
Degree:
Master
University:
University of Baghdad - Faculty Of Administration And Economics - Department Of Statistics
Language:
Arabic
University location:
Baghdad
First pages:
07T3272 - p.pdf
Abstract:
The basic idea of the search started from the basis of operations research applications in the study and assess the reliability of producing machines that require decisions assigned to the foundations of modern scientific, And the evolution of technology led to the use of reliability index in the studies of design, operation and maintenance, and increased interest in the field of plants in general and chemical plants in particular.On this basis, has become a study and evaluate the reliability of machines and equipment as part of the effective jurisdiction of the Operations Research.Therefore, the constant interruptions in production machines gratings batteries in the General Company for the manufacture of liquid batteries (Babylon Lab - 1 - ) in Iraq, And the absence of estimating method dependents Parametric methods and Nonparametric ofthese machines led to a lack of knowledge stop hours of work these machines, What are the best methods that can be adopted in the estimate and therefore did not fulfill the above obligations the company in achieving energy production.So work requires an estimate of the dependents of these machines for the purpose of access to identify and stop engaging in hours and therefore maintenance programming and production of the above.Based on the foregoing, this research included mainly a comparison to estimate the reliability function in two approaches , first approach methods and parametric estimation methods, which included the following : 1 - Weibull Probability Plotting Method (WPPM).2 - Maximum Likelihood Method (MLM).3 - Method of Moments (MOM).4 - Rank Regression on Y Method ( RRYM ).5 - Rank Regression on X Method (RRXM).6 - Nonlinear Weighted Least Squares Using The Empirical Distribution Function Method(NWLSM).Either the second approach Nonparametric methods, which included the followingmethods : 1 - Empirical Methods (EM).2 - Product Limit Estimator Method(PLEM).3 - The Empirical Kaplan Meier Estimator Method(EKMEM).4 - The Empirical Weighted Kaplan - Meier Estimator Method(WEKM).5 - The Modified Kaplan - Meier Estimator Method(MKMM).6 - A weighted Estimation Method for reliability function(WMR).7 - Modified Method One(MMO).8 - Modified Method Two(MMT).The research aims to hold an estimate of the reliability function by Parametric methods for the distribution (Weibull with two parameters ? and ?), as well as estimation reliability function by nonparametric methods and compare the estimators of the reliability function by parametric and nonparametric methods to reach to the best parametric and nonparametric method, Through reliance on the index, the statistical integral mean square error (IMSE), And using the simulation and volumes of different samples as an estimate and is designed for dependents of machines that will allow the Department of gratings produce batteries out hours of operation of these machines and duration of stops and the application of parametric and nonparametric method and the best on the real data (failure times) Submitted by the pilot.The researcher has proposed two modified approach (MMO&MMT) in which Proposed modified weight to the empirical weighted kaplan - meier estimator has surpassed two approach, Which gives less integral mean square error (IMSE) of the integral mean square error to the nonparametric approach (The Empirical Weighted Kaplan - Meier estimator method (WEKM)). The researcher reached several conclusions, including that the best parametric method to estimate reliability function is a (Maximum Likelihood MLM) and for the best nonparametric method is a (Empirical Methods (EM)), and relying on standard statistical integral mean square error integral (IMSE).
