應用類神經網路技術分析產品報價系統之研究-以運動器材業為例

The Application of Neural Network System on Product Quotation System – Example of Fitness industry

黃建龍(Chien-lung Huang)

指導教授 ： 李見發 ； 張阜民

朝陽科技大學/管理學院/財務金融系/碩士(2010年)

摘要

由於近年來產業供應鏈國際化，與全球新興代工市場崛起的雙重作用，台灣地區產業競爭力正快速的凋零。從成本結構中的原物料價格、加工成本、固定管銷費用等三大類項分析。近年來台灣已經失去了低價原物料和低廉人工的優勢，因此為了要在商業的競爭中生存，如何提高產品品質，和快速、彈性的價格策略，是一個值得研究的課題。市場上不變的競爭要因是『價格』，然而定價策略必需用產品真實成本做基礎。原物料價格的波動在影響報價比重上，均遠大於其他二項，所以確實掌握原物料價格的波動，也就直接控制產品成本，和市場競爭力的關鍵。
本研究首先利用傳統產業中現有的電腦作業系統，讓不同作業系統中各不相同形式的資料流，交互共用，提升企業效率。進而建立一套適合產業使用之標準化作業程序。由於目前尚未有相關使用類神經網路技術分析物料價格之研究。本文將透過類神經網路的技術，對原物料價格進行分析，進而建立產品報價預測，期使企業在推展營運決策時，有效的規避風險，及策略運作提升傳統產業競爭力。

關鍵字

企業資源管理 ； 產品資料管理 ； 類神經網路

並列摘要

No matter how the market has evolved in the recent years, traditional industries have remained to be the most important foundation of this nation. Without the support of this foundation, most of the advanced high-tech industry achievements could have not been realized today. However, the traditional industries in Taiwan are facing crisis in the competition against globalization and challenged by OEM factories from overseas. The analysis on the production capital cost concerning the production material, human resource and fixed management expenses has revealed that cheap production material and human resource is no longer the advantage in the traditional industries of Taiwan. In order to survive within the market competition, the ability of producing high quality product with an efficient and flexible production system has become a key issue.
The most important competition factor in the market has always been the price of product, which is influenced by the cost of product capital. Out of the three expenses mentioned above that production material expense has showed the most violent volatility. Since production material expense has also constituted the largest proportion of production capital, controlling production material expense volatility has become the key factor of controlling production capital cost and the enhancement of market competitive ability.
This study has focused on the co-operations of computer programs currently adopted by the traditional industries with different functions to inter-communicate and share database. This paper intends to establish an ideal working program for the traditional industries to standardize product pricing and evaluation process. An artificial intelligence program based on Neuron Network System is also adopted in this research to simulate production material cost and co-operate with the product pricing and evaluation system to set up a product capital cost forecasting system. This forecasting system shall assist company CEO in the daily operation decisions with better efficiency and flexibility and reducing operation risks.

並列關鍵字

PDM ； ERP ； Artificial Neural Network

參考文獻

[11]莊富鈞（2002）。以BOM為基之二階供應鏈生產環境分析。碩士論文，中原大學工業工程研究所，中壢。

[1]李惠妍（2002）。類神經網路與迴歸模式在台股指數期貨預測之研究。碩士論文，國立成功大學企業管理學系，台南。

[3]林建君（2005）。ERP系統參數設定之導入分析-以個案食品公司生產製造系統之損耗參數為例。碩士論文，國立中央大學企業管理學系，中壢。

[14]黃明傑（2008）。伺服器物料架構管理中PDM之專案知識整合。碩士論文，元智大學工業工程與管理學系，中壢。

[1]Chou, Y.J., C.F. Ouyang, and W.L. Kuo, (2003). “Feedforward Artificial Neural Networksfor Nutrients Removal Simulation in Multiple Stages Enhanced BiologicalNutrient Removal Process,” Journal of the Chinese Institute of Engineers, Vol.26,No. 2,pp. 211-219.