Based On Improved RBF Transmission Line Project Investment Valuation Optimization Research

With the development of market economy, especially under the State GridCorporation large construction management system standardization andspecialization goal lead, power grid construction projects cost control level request toimprove, none of the traditional fixed budget system has been unable to fully meetthe needs of the construction, it is necessary intelligence model which widely usehistorical data to estimate the cost of the investment project to solve this problem.The last two decades, fuzzy mathematical algorithms, linear regression, gray theoryestimation method, a variety of intelligent algorithms are successively applied to thereal estate and construction projects such as road and bridge construction costforecast. But relatively speaking, fuzzy and gray theory established model andalgorithm design is relatively simple, Ignore factors and assumptions are more. TheBP neural network involved in the convergence of the algorithm, robustness andgeneralization is poor. Based on improved genetic algorithm RBF neural networktransmission line project investment valuation studies put forward an intelligentimproved learning algorithm with genetic algorithm that optimization RBF weights,And the algorithm is applied to the transmission line project investment valuation.Paper introduces the basic theory of artificial intelligence, neural networks andgenetic algorithms theory on the one hand, based on the existing valuation modelscommonly used neural network carried out a detailed analysis and research, thenfurther proposed RBF network valuation model based on genetic algorithmoptimization to minimize the influence of subjective factors on the outcome, full useof genetic algorithms’ global search features to self-correction width of the RBFnetwork, the center and the hidden layer weights, thereby greatly improving theaccuracy of the model results.On the other hand, according to the collected110kvtransmission line project investment cost data, use the SPSS component analysis toextract the main factors that influence the engineering project cost as the model inputfeature vectors, and use MATLAB as a platform for simulation analysis, shows thatthis model is not only calculation simple,fast, and has higher calculation accuracy.Finally put forward the method not only can validated the rationality of projectcost analysis,also can undertake reasonable for new project, can also achieve rapid price in the bidding, it has good application value in the field of engineeringconstruction. At the same time, in the end of the article summary and outlook the fulltext and point out the direction of further research.