Applied Research On The Model Of Artificial Neural Network To Quickly Estimate The Project Cost Of Highway

The highway is an important part of the development of China’s transportation construction, it is related to China’s rapid economic development and the improvement of people’s living standards. In recent years, China’s highway construction has developed rapidly, increasing the requirements of the building of the highway, the gradual increase in the investment in highway construction. However, with the investment continued to increase, the cost of out-of-control phenomenon is more and more serious, especially highway construction project cost subject to a variety of factors, the composition is more complex, more cumbersome calculation vagueness in the investment process large extent a waste of money. The main reasons contributing to this phenomenon is the pre-construction did not cost an accurate estimate of the overall functioning of the entire project cost of preliminary engineering is closely linked. Due to shortage of funds, the construction unit to do the construction of highways, these waste must be strictly controlled, in the project prior to the implementation of more accurate estimate of the project cost to reduce or even solve such waste a kinds of practical methods, in order to achieve the purpose of the control project. Therefore, it is possible to accurately estimate preliminary cost is very critical, but the application cost estimation method is not very scientific, which requires the establishment of a way to quickly and accurately estimate the highway project cost.This paper analyzes the pros and cons of the existing estimation method, based on the characteristics of the highway project cost, fast highway construction cost estimates based on artificial neural networks, a specific study to make a quick estimate of the aspects of the application of the highway project cost. Specific screening basic knowledge of the highway’s construction cost data collected Shandong territory few have been built using fuzzy math, and engineering data as training samples selected to be estimated project close to take theoretieally more mature improvement BP neural network algorithm model. To verify the accuracy and feasibility of the model created by the contrast clause highway actual and predicted values, and provide a new method for the calculation of the project cost.