الفهرس 
Preliminaries and Methodology
Spare Parts ClassificationOnly 14 pages are availabe for public view
 |  | from | 235 |  |  |
| | | from | 235 | | |
Abstract
Due to the operating conditions, products in service are subject to failure that resulted from damage to their parts. Therefore, the production plan is needed to provide the needs of service parts. The production plan should be placed based on an accurate forecast technique to predict the required demand for service parts. There are many costs concerned with service parts such as production cost and inventory costs. So, a precise forecasting technique for saving the required service parts with suitable ranges is essential for minimizing the penalties resulting from over or underestimation of demand. The required demand for service parts is controlled by many parameters related to the part itself and the product that contains this part. These parameters differ from place to place and in time. Therefore,service parts having a stochastic nature are difficult to predict easily. The forecasting process should take all changes in all parameters in the product reliability and the failure of its parts into account. In this study, fuzzy logic is integrated with the proposed forecasting probability technique to predict accurately the required demand for service parts based on a proposed production policy. The fuzzy concept is applied by using alpha cut and triangular fuzzy numbers. For the defuzzy process, four different methods such as mean-max, centroid, signed distance, and graded mean integration representation methods are used. The proposed forecast model is controlled by the part failure rate and the reliability of the product. Therefore, the study is introduced in two states: crisp state, and fuzzy state. The fuzzy state is presented in three main cases to study the sensitivity effect of the fuzzy parameter on the forecast demand. The first: is the forecasting of service parts based on the fuzzy reliability of the product. The second: is the forecasting of service parts based on the fuzzy failure rate of the part. The third: is the forecasting of service parts based on the fuzzy of both the reliability of the product and the failure rate of its parts. The periodical deviation percentage is calculated between the crisp and each fuzzy case. from the suggested fuzzy cases, each fuzzy case gives a different deviation percentage. For important industries that need to provide service parts permanently whatever their cost, it is preferred to select the maximum deviation percentage. Otherwise, select the minimum deviation percentage. In this study, the effect of the fuzzy mean of the part failure rate has a maximum deviation percentage of 12.768%. This is a suitable range of forecast demand that would consider reasonable for saving stock from running out and avoid excessive costs resulting from under or over-estimation of demand.