Research On The Knowledge Base Of Geological Alignment Selection Cases Based On CBR

For highway(railway)geological route selection,design methods rely on expert experience and manual decision-making,and geological knowledge is difficult to in-depth route selection process.The knowledge of typical geological route selection examples is scattered in various documents,design manuals and other related materials,and it is difficult to provide sufficient experience support for alignment designers;BIM(CAD)technology fails to integrate GIS and manual in the process of route selection design.Smart and other means.Based on this,the thesis aims to establish a knowledge base of geological route selection examples,using technical methods such as case-based reasoning(CBR)and spatial relational reasoning,to establish a knowledge base reasoning mechanism,and to integrate with GIS and BIM(CAD)systems to realize geological analysis.Decision information support for the route selection design process.The thesis fully considers the knowledge,spatial and interactive characteristics of geological route selection in the design process,puts forward the design idea of BIM(CAD)+GIS,and establishes knowledge-based reasoning+BIM(CAD)interactive design+GIS spatial reasoning System architecture.Taking karst geological phenomenon as an example,the thesis has collected a large number of examples of geological route selection.According to the characteristics of geological knowledge representation,the geological route selection examples are divided into problem domains containing engineering background and geological overview,and solution domains containing alignment design schemes and engineering pictures.The problem domain representation method based on attribute-value pairs is proposed as the reasoning basis of the case knowledge base.Aiming at the characteristics of strong subjective color and incomplete information in expert scores,a method based on machine learning text mining technology to extract keywords to determine feature attributes is proposed,and then an attribute reduction algorithm based on semantic similarity is designed to determine the geological route selection.The storage structure of the instance knowledge base.In view of the difference in the information requirements of the instance knowledge base in the design process,the paper establishes a reasoning mechanism based on attribute similarity,adopts the method of incomplete additive language judgment matrix to determine the attribute weight,and realizes the relationship between attribute values through the deep learning algorithm.Correlation calculation.Aiming at the spatial relationship between geological phenomena and alignment objects,a method of describing the spatial relationship between alignment and geological objects using the dimensional expansion nine-intersection model is proposed,and a digital geological object model is established,which is based on spatial relations+attribute similarity.Degree of reasoning mechanism.The thesis uses Autodesk Civil 3D+Map 3D as the platform to solve the key technical problems of knowledge and design information mutual feed mechanism,BIM(CAD)user interaction mechanism,BIM(CAD)and GIS information transformation mechanism,and complete the knowledge of geological route selection examples.The development of the database system provides decision-making information support in the design process of geological route selection,and verifies the operability of the geological route selection instance knowledge base system with specific examples.