Design And Implementation Of Machine Resource Management System

ERP (Enterprise Resource Plan), which is a very popular system on managing company resource, is growing to be an advanced ideology on company management field. However, there are lots of adverse factors which would lead a failure on importing ERP software. One of the criticality factors is the design of software itself. In this document, we focuse on designing a scientific solution to improve the robustness of the software through studying the idea of ERP development. Thus, we can reduce the risk of bring in an ERP software, lower the cost of software maintenance and upgrade.In the document, we take machine resource management system as an instance. First, we start with analyzing the requirements of system, and dispart it into four major functions:basic information maintenance, host assignment, forecast function and report function. Inside the basic information maintenance function, we have three sub-functions:role control function, project maintenance, machine maintenance. Also beside the function requirements, there is non-functions need to be satisfied, like system performance requirement, security requirement, usability requirement and so on.While achieving of the functions, we append arithmetices on host assignment function and forecast function. In assignment function, first we use the Banker's algorithm to calculate the safe secquence, and then combine the project level and peoject length to generate a new project priority, using which we can get the optimal secquence from safe secquence, and then assign hosts by the optimal secquence. In forecast arithmetice, we take Grey prediction as basic theory, and built GM (1,1) model to analyze the history data. By calculating the fitted value of data secquences, we forecast the data of future. Also, we apply differential analysis function to make sure the forecast data accurate.We use B\S to implement the whole system, using MVC patten to construct the frame of system. And by enhancing the frame classes, we reduce the coupling of codes and avoid code reuse. Using JavaScript to describe interfaces, and Ajax technique to handle asynchronous data transmission, which makes good system performance. In data layer, we use store procedure and SQL scripts as main methods to handle data problem.