Research on Construction Cost Control Based upon BP Neural Network and Theory of Constraint

Abstract

Construction industry is an important material production sectors of the national economy, it has been an important pillar industry in China's economic development. Though the subprime mortgage crisis in the U.S. in 2008 has brought great impact on the global economy, large numbers of strong measures of domestic demand expansion issued China government which promoted stable and rapid economic growth and obtained remarkable achievements. A major reason for these achievements is a lot of construction investment, so the construction cost control is especially important when a huge capital is invested. However, In China, the current construction cost pricing model is changing from quota valuation pricing to bill of quantities (BOQ), the root of getting out of control is derived from the defect of the construction cost pricing model. This paper applies BP neural network to forecast the construction cost, and uses the outcome of the forecast to optimize the bill of quantities (BOQ), then takes advantage of the theory of constraint to control the construction cost based upon optimized the BOQ at last. A sample of data from six highways, including a testing data, using BP neural network simulation, building simulation model., The model fit well through 13 iterations, and the outputs of test data is consistent with the actual. Then using TOC to optimize based on the output, compared to the previous final account, the optimized final account of the project represents a reduction of 20.726 million yuan, lower by 1%.