基于PLM的ARIMA决策支持模型及应用

发布时间：2018-11-23 06:38

【摘要】：随着经济社会的转型升级,制造型企业的竞争加剧,企业在生产计划制定、原材料采购和产品销售等方面需要做到更加精细化和规范化,需要对企业PLM过程中的产品设计、制造、装配、入库、出库、销售、退货及服务等各环节的变化情况进行实时监测,以过往的企业生产数据为样本和基础对企业未来一段时间内的产品生产、销售等过程进行预测,为企业制定采购计划、生产计划等战略决策提供数据的支撑。本文首先分析了企业PLM过程中决策支持的重要意义以及数据预测对企业决策的影响,以企业过往生产数据为研究对象,对如何获得有效的预测数据以及以友好精美的方式展示给决策者进行数据预测的建模和应用研究,具体的研究内容如下:首先,构建了基于ARIMA的数据学习预测模型,采用残差检验对模型进行校验,并用方差校验和T校验对模型进行优化处理;接着,采用R语言和基于Java的JSP、Servlet和JavaBean技术以及rJava相结合的方法,实现预警模型求解并进行测试验证与优化;然后,构建了基于浏览器/服务器结构的决策支持系统,在为企业决策提供支持的同时根据不断变动的企业生成情况对预测模型进行微调;最后,以决策支持系统为基础建立基于web的应用,让决策者可以及时直观的了解企业实时的生产情况。
[Abstract]:With the transformation and upgrading of economy and society, the competition of manufacturing enterprises intensifies. Enterprises need to be more refined and standardized in production planning, raw material procurement and product sales, and need to design products in the process of enterprise PLM. The changes in manufacturing, assembly, warehousing, delivery, sales, return and service are monitored in real time. The production of products in the future will be monitored on the basis and sample of past enterprise production data. Forecast the process of sales and provide data support for strategic decision-making such as purchasing plan and production plan. This paper first analyzes the importance of decision support in PLM process and the influence of data prediction on enterprise decision making. The research on how to obtain effective prediction data and display it to decision makers in a friendly and exquisite way is as follows: firstly, a data learning prediction model based on ARIMA is constructed. The model is calibrated by residual test and optimized by variance check and T check. Then, R language, JSP,Servlet and JavaBean technology based on Java and rJava are adopted to solve the early warning model and test verification and optimization. Then, a decision support system based on browser / server structure is constructed, which can provide support for enterprise decision and fine-tune the prediction model according to the changing situation of enterprise generation. Finally, based on the decision support system (DSS), an application based on web is established, so that the decision-makers can understand the real-time production situation of the enterprise in time and intuitively.
【学位授予单位】：贵州师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F426.48;TP311.13

【参考文献】