Font Size: a A A

The Research On Cost Estimate Of Constructional Engineering Based On Artificial Intelligence Technology

Posted on:2010-04-23Degree:MasterType:Thesis
Country:ChinaCandidate:Y XiongFull Text:PDF
GTID:2132360278962299Subject:Structural engineering
Abstract/Summary:PDF Full Text Request
To determine the engineering construction's cost is one of important contents in the engineering cost management, especially in the earlier stage stage of engineering project, on the premise that construction's information know less,insufficiently, estimation,budgetary estimate have shortcomings that are big error and waste a lot of time in formation. So to take Jiangxi Province as the example, collect and organize the material of engineering cost , establish preliminarily the Jiangxi Province's engineering cost database; on how to establish a estimate model of a higher computational precision and quicker computational speed has made the detailed research and the discussion.The paper's primary coverage has included two aspects,one is to introduce the artificial intelligence technology, the neural network theory and the genetic algorithm , then analyses and studies deeply the estimate model of the existing commonly used neural network., point out neural network structure's own flaw and perfect generalization ,reduce the subjective factor to influence the result as much as possible, proposes the estimate model bases on the genetic algorithm optimized the RBF network , make full use of the global search of genetic algorithm, let the RBF network's width, center and threshold of the concealment level simultaneously to self-revision through the genetic algorithm to enhance greatly the model's the computational precision.The other is to indicate the primary factor of influencing engineering project cost currently, according to have collected date of engineering construction cost,analyze that how much the cost have been influenced by the characteristic variable on the basis of having proposed the model of genetic algorithm optimized the RBF network and the most sensitive characteristic variable will become the model's input variable (characteristic variable), simultaneously have also proven those characteristic variable have the intense relevance with cost. Finally, we establish preliminarily database of Jiangxi Province engineering construction cost based on frame construction through classification, reorganization and formulation, and take MATLAB as the platform, has compiled the program, through the example computation, confirmed this model not only calculates more easily and quickly but also have more accurate outcome.At last, we have carried on the summary to the paper's work and pointed out the future possible research direction and work.
Keywords/Search Tags:engineering cost, estimate model, neural network, genetic algorithm, sample
PDF Full Text Request
Related items