Competitive power of an enterprise depends primarily on the technology, resources and management capability being possessed by the enterprise, and such competitive power eventually materializes in the form of competitive price of the enterprise. Judging from the characteristics of the engineering industry, every engineering company have to cope with other engineering companies to garner the projects in the markets of competitive open bidding system under the equal opportunity, and in this regard, the cost estimation system takes greater part of the role in the marketing business of the engineering industry especially to maintain the competitiveness in the price.
The problem how to improve and enhance the reliability and quality of the Plant Cost Estimation System is the one of prominent concerns of engineering companies. However, no particular solutions or researches has been publicly known so far, and this is the actuality of engineering industry of our country. This study elaborates to propose the introduction of Expert System with a view to improve the quality of the cost estimations system and suggest the practical application of the system. An actual example of the system application is reviewed as well.
This study on the practical application of the Expert System to the current plant cost estimation system will bring on such results as: Firstly, a decision tree will be generated to extract the decision rules for cost evaluation by analyzing main variables affecting the cost, and a cost evaluation matrix will be provided to facilitate the use of them as a scale of cost decision. Secondly, cost prediction will be figured out in real time to prove the possibility of building-up the efficient and prompt Plant Cost Estimation System in an improved manner. Finally, integrated application of the Expert Cost Estimation System will be made to effectuate the epoch-making development and renovation of the cost estimation system as well as a kind of Decision Support System in Engineering Company.
In the course of this study, C4.5 Program of Inductive Learning System and BPN5 Program of Neural Network System have been used.