售后服务数据的运用模型

上传者：网友
|

翻新时间：2022-12-12

收藏到账号

全屏

售后服务数据的运用模型

摘要

本文以工厂提供的轿车某部件的千车故障数的数据表为研究对象,在分析了数据表的特征并剔除了表中不合理的数据后，根据数据的不同特点建立了多个模型，对其中3个生产批次的部件的质量状况进行了预测，探讨了如何改进数据表以进1步提高数据的有效性，并对厂家的质量管理提出了1系列建设性意见。

本文的主要工作有:

1对数据进行了分析，提出了原文中千车故障数的定义存在的几种不合理性,并对其进行了修正,给出了更加合理的千车故障数的概念；

2 采用横向最小2乘拟合与纵向卡尔曼滤波方法的联合预测方法对数据表进行填充；预测结果为:

0205批次使用月数18时的千车故障数为 79.65；

0306批次使用月数9时的千车故障数为 32.78；关键词：最小2乘法；灰色预测；马尔柯夫链

Post-sale service data utilization model

Abstract

This paper factories car a 1000 car parts fault of a few data tables for study , After has analyzed the data sheet characteristic and has rejected in the table the unreasonable data, Has established many models according to the data different characteristic, Has carried on the forecast to 3 productions raid of part quality condition, How discussed improved the data sheet to further to enhance the data the validity, And gave a series of constructive comment to the factory quality control.

The main work:

1 on the data analysis, The original 1000 car fault the definition of several unreasonable, And its amendments, Is a more reasonable number of 1,000 cars the concept of fault.

2 horizontal and vertical least squares fitting method of Kalman filtering method of predicting the data sheet filler Forecast results:

0,205 batches of 18 for the use of 1,000 cars at 79.65 Fault 0,310 installment on the use of several 1000 12:00 Trouble at 12.57

4 after-use method to predict the test data

5 Finally, the re-tabulation of the proposals

Keywords: least squares; Gray forecast; markov