基于人工神经网络的非标产品制造成本估算研究  

Research on Artificial Neural Networks for the Manufacturing Cost Estimation of Non-standard Product

作　　者：詹月林[1] 

机构地区：[1]盐城工学院,盐城224051

出　　处：《价值工程》2014年第14期203-205,共3页Value Engineering

基　　金：江苏省高校自然科学研究项目资助(10KJB480003)

摘　　要：大型和形状复杂的非标产品的生产成本,是很难估算的。目前采用的人工方法估算的误差较大。当成本估算用于招标目的时,这样的误差将产生不利的后果。成本高估承担的风险,是使公司缺乏竞争力,失去客户,而低估了成本,赢得了合同,但招致财务的损失。因此,一个精确的产品生产成本估算方法,对项目管理而言至关重要。本文应用神经网络建立的参数模型,针对大型和形状复杂的非标产品的生产成本估算和目前采用的手动估算相比,精确度有了很大提高。Estimating the manufacturing cost of large-sized and complex-shaped non-standard product is hard. Now the manual estimation methods suffered from a high estimation error. When cost estimates are to be used for bidding purposes, a poor accuracy may have detrimental financial consequences. A cost overestimation bears the risk of making the firm uncompetitive and losing a customer, while underestimating the cost leads to winning a contract but incurring a financial loss. Therefore, a precise cost estimation is critical for project management. In this paper, building the parametric model of artificial neural networks, compare with the artificial neural networks and the manual estimation methods in estimating the manufacturing cost of large-sized and complex-shaped non-standard product, the accuracy is improved.

关 键 词：神经网络(ANN) 成本估算 非标产品 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]