Research And Implementation Of Electric Power Engineering Cost Budget System Based On CBR

As the pillar of national economy,infrastructure construction needs to be developed first.The construction of large-scale and high-quality power projects is a prerequisite for infrastructure construction.However,the cost management of power engineering in China lags behind the international level,which restricts the improvement of the quality of domestic power engineering to a certain extent.For a long time,a large number of historical data has been produced in power engineering project,which contains a wealth knowledge of construction cost.Therefore,how to excavate effective information from the past project cost cases,and help improve the project cost management,has become an urgent problem to be solved.In recent years,the method of engineering cost estimation has been a research hotspot,and some existing methods can complete the tasks of engineering cost estimation in particular fields.However,there is still a lack of a power engineering cost model that can continuously learn cost knowledge from historical data.In order to make up the gap,this paper summarizes the research progress of engineering cost estimation methods and casebased reasoning and related technologies at home and abroad,and determines that the specific task of power engineering cost estimation is how to use historical power engineering cost cases to build a new power engineering case.In this paper,based on case-based reasoning technology and deep learning technology,the cost estimation model of power engineering——PES is established by using the cost data of historical power engineering projects.The model has two key stages: in the retrieval stage,we learn the similarity differences in old cases and recommend the adaptive cases with the highest similarity to the users;In the prediction stage,the solution of the adaptation case is modified accordingly with the adaptation knowledge obtained,so as to solve the new case.Specifically,first of all,by using multidimensional scaling change algorithm and kmeans clustering algorithm to complete the feature dimensionality and clustering and obtain the optimum clustering five groups,then selecting and extracting the feature and processing the data according to the optimal cluster groupings of original data,the power engineering cost estimation data set was obtained.Second,in the PES,first put forward the joint similarity calculation method fusing power engineering ontology similarity and power engineering comprehensive feature similarity,and build a case base according to the joint similarity calculation method,further,by using artificial neural network to learn power the implicit expression between engineering case characteristics and cost of static and dynamic.Subsequently,the retrieval results and prediction results are fused to generate the proposed scheme.In the scheme revision stage,the scheme was revised and evaluated by the expert is the final scheme.To sum up,this paper finally designed and realized the power engineering cost budgeting system based on case-based reasoning.