The Study Of Two-dimensional Warranty Data Analysis And Flexible Preventive Maintenance Strategy Design

Great changes have taken place in information and communication technologies with deepening economy globalization.Beyond price reduction,high quality and excellence performance,attractive post-sale service has played a key role in encouraging market demand.In this context,most durable products are sold with warranty policies.According to the current research progress of product warranty in manufacturing and present problems in the product warranty management,this dissertation discusses methods to evaluate the field reliability,builds up warranty claim prediction models and determines the optimal warranty policy for products sold with a two-dimensional warranty.The major results as listed as follows:Firstly,this study proposes a method to model field reliability based on two-dimensional warranty data with censoring times.The usage rate is treated as a random variable in the analysis of warranty data.Both the claims data and maintenance data are used in formulating usage rate distribution,which can provide manufacturers with the information on customer behavior.Since it is costly to obtain the information of censored data,the stochastic expectation-maximization algorithm is chosen to estimate parameters of the field reliability model.Both extensive simulation studies and real warranty data collected from an automobile manufacturer in China are conducted to test the performance of the proposed model.The results show that the model built with the stochastic expectation-maximization estimation provides more accurate reliability estimation than that based on the maximum likelihood estimation.Secondly,seasonal index model is adopted according the phenomenon of the number of claims with a two-dimensional warranty showing seasonal fluctuations.Firstly,the usage rate is considered in calculating the failure rate function.Then,seasonal index model is introduced to model claims with a two-dimensional warranty.The results of case study show that seasonal index model outperforms the traditional model,such as the non-homogeneous Poisson Process and the autoregressive integrated moving average model.Thirdly,we present a warranty claims forecasting model for new products sold with a two-dimensional warranty.In the case of new products,an important feature is that the new generation of products has a close connection with the previous generations of products it replaces.Thus,the warranty~I ~Iclaims data of the previous generations of products can be used for building a warranty claims forecasting model of new products.The effect of usage rate on product degradation is considered.Then the non-homogeneous Poisson process is used to model failure counts of repairable products and the constrained maximum likelihood estimation method is used to estimate model parameters.The results of data experiments based on both simulation and real data collected from an automobile manufacturer in China show that the proposed model considering the varying usage rate outperforms the traditional models in forecasting the number of warranty claims.Fourth,we consider the problem of flexible periodic preventive maintenance planning under a two-dimensional warranty policy for the repairable product.Considering differenty types of customers,we develop a mathematical optimization model based on flexible preventive maintenance,in which the number and degrees of preventive maintenance are decision variables and maximizing the profits of the manufacturer is the objective function.Besides,the effect of the warranty proportion,cost parameters and the warranty coverage on the optimization results are investigated.The provided results reveal that the two-dimensional warranty policy with flexible preventive maintenance strategy could not only effectively improve the profit of the manufacturer,but also provide guidance for warranty policy design.