การประมาณราคาค่าก่อสร้างสำหรับอาคารงานราชการโดยใช้เทคนิคการสร้างแบบจำลองพยากรณ์

Abstract

Due to the overall contraction of the construction industry in the country, the bidding competition required effective precision to prevent losses from too-low price in the bidding process, either in the private or public sector. However, before the bidding process, the bidders must have an estimate of the construction cost before the bidding. There are two widely used methods for construction cost estimation as a rough estimation, which has the advantage of being able to estimate construction cost quickly but has the disadvantage which is a high price tolerance, and a detailed estimation, which its advantage is the ability to estimate the construction costs more accurately, but its disadvantages are the need for a complete construction plan and a more time-consuming process. After considering the disadvantages of the 2 construction cost estimation methods, research for modeling the government construction cost estimation was conducted by using the technique of the forecasting model, the econometric model by using the linear regression analysis technique and the artificial intelligence model by using the Artificial Neural Network technique. There were 11 variables in total as 1) usage area of the building 2) average perimeter 3) average inter-floor height 4) building height 5) the number of floors 6) roof area 7) bathroom area 8) ground slab 9) open space 10) type of roofing material and 11) type of floor structure. For the result of linear regression analysis technique, the study showed that R2 = 0.732. Plus, the study result of Artificial Neural Network technique showed that the model consisted of 2 hidden layers which each layer has the number of nodes at 10 and 8 nodes, respectively, with the best Root Mean Square Error value at ± 0.331 million Baht. When the new data set was tested for validity, it was found that R2 = 0.914 which has proved the accuracy of the forecasting model as an alternative for government bidding participants to reduce the tolerances and to spend less time to estimate construction costs more efficiently.