Research On Evaluation Model Of Railway Engineering Geological Conditions Based On Big Data Clustering Mining

Sichuan Tibet railway is a very important construction project in China’s 13 th five year plan,and its construction has been widely concerned by compatriots of all ethnic groups.In the design,construction and operation stages of railway engineering,it is necessary to investigate,analyze and evaluate the geological conditions along the railway engineering to ensure the engineering safety.The evaluation of railway engineering geological conditions based on big data clustering mining plays an important role in improving the efficiency,objectivity and comprehensiveness of the evaluation results.The traditional evaluation method of railway engineering geological conditions has some shortcomings,such as strong subjectivity,low evaluation efficiency and non intuitive evaluation results.In view of the above limitations,this paper proposes an evaluation system of railway engineering geological conditions based on the risk of geological disasters,uses relevant geological disaster influencing factors for big data mining,establishes an evaluation model,and carries out big data visualization display.1.Use crawler,httprequest and other data acquisition technology to collect and preprocess the data of geological hazard risk factors.At the same time,research and build a big data processing platform based on Hadoop + spark framework to provide efficient data access performance,and provide efficient computing performance for big data analysis and mining.It realizes the efficient and accurate automatic collection,preprocessing,storage and calculation of multi-source heterogeneous big data.2.Based on the risk of geological disasters,this paper analyzes the big data of railway engineering geological conditions,analyzes the correlation between railway engineering geological conditions and various influencing factors,and provides theoretical support for the follow-up big data mining.3.Research and implement the feature selection method based on random forest,and use the random forest algorithm to establish the classification model of the original training set.The classification model is optimized,and the weight value of each characteristic attribute is output.Combined with the weight value and the conclusion of big data analysis,the characteristic attribute selection of the training set of railway engineering geological condition evaluation model is completed.4.The concept of difference weight density is proposed,and it is introduced into the selection of initial clustering centers of K-means clustering mining algorithm,and an improved algorithm MDDK means is proposed.The algorithm overcomes the randomness of the initial clustering center selection of K-means algorithm,and improves the accuracy of clustering and the efficiency of execution.Compared with traditional K-means algorithm and density based k-means algorithm,the experimental results show that the improved algorithm has higher clustering accuracy and efficiency.5.The evaluation model of railway engineering geological conditions based on MDDK means algorithm is proposed.The model evaluates the risk of regional geological disasters based on the data of geological disaster risk influencing factors along the railway,evaluates the railway engineering quality conditions in the region,and applies the model to Sichuan Tibet railway.