Research And Application Of Prognostics And Health Management(PHM) In ERP System

After two decades of information technology popularization and promotion, currencies, the value of management software has been widely recognized in Domestic enterprises. Especially with the ERP (Enterprise Resource Planning) software development and application technology matures, more and more enterprises pairs have a better understanding of the ERP. ERP software is used more and more widely. ERP software is a complex software system, the anomalies or failures consisting of hardware, operating system, network, database, middleware may lead to abnormal working in turn affect the normal operation of enterprises or even interrupted. Therefore, how to monitor and manage a healthy ERP system and try to early detection fault and response to failure are of great value.PHM(Prognostics and Health Management) technology focuses on health management, fault prediction, protection as appropriate. It has been widely studied and used in military equipment and complex important mining equipment protects areas of working. This thesis intends to research PHM techniques used in the field of the safeguards ERP system, in particular implementation and verification of the important technical aspects:data collection, data processing, data prediction.The main work completed is as follows:1. For method of ERP system health management, this thesis summarized the current status of such products, analyzed the management needs and problems.2. Based on the core of PHM technology and the actual usage in other areas, the author analyzed and estimated the realizability of core elements:data collection, data processing, data prediction and data fusion in the field of software systems.3. According to the characteristics of the ERP system, SNMP (Simple Network Management Protocol) was selected as the main method of data collection. The author has completed the program for data-gather using the JAVA programming language. It laid the foundation for the subsequent software development. 4. Using the basic Gray prediction model, new dimension and Markov chain, the author presented a more accurate method to predict ERP system performance data. By using A rapid prototype testing by MATLAB tools, the rapid build of software development was studied.5. For fusion of multi-type data, the D-S (Dempster-Shafer’s) theory be chosen to attempt to establish the mapping from the running status data to the fault phenomenon.This thesis introduced the PHM technology for the software systems of the health management, analyzed and validated the feasibility of the core technology combining with the actual work. The design of the PHM software architecture and main components function laid a solid foundation for the realization of specific software.