Blasting vibration prediction method based on Spark gene expression optimization

Abstract

The invention provides a blasting vibration prediction method based on Spark gene expression optimization, and involves the field of machine learning. The data preprocessing technology is used for processing blasting data, and a sample data set is obtained; by calculating the incoordination rate of each condition attribute, condition attributes with the incoordination rate lower than a threshold value are deleted from an original data set, and then a new data set is generated to serve as an input data set; on each node, an improved gene expression method is used for function optimization, a blasting vibration effect prediction function can be obtained, and then the prediction value of the blasting vibration peak velocity is obtained. The method can better solve the training efficiency problem under the condition of mass data, the new-generation parallel computing technology is adopted, the genetic expression programming algorithm is improved for the global function optimization of theblasting data, the convergence speed can be largely improved, and the training efficiency is improved without affecting the training precision.

Description

technical field [0001] The invention relates to the field of machine learning, in particular to a blasting vibration prediction method. Background technique [0002] Blasting vibration usually refers to the strength of the vibration wave generated by blasting, which can be described by the speed, frequency and time of blasting vibration. In practice, the peak value of blasting vibration velocity is often used to describe blasting vibration. Blasting vibration prediction is to study the relationship between blasting vibration influencing factors and blasting vibration, among which blasting vibration influencing factors include total charge, segment charge, blast center distance, geological conditions and elevation difference and many other factors. [0003] Prediction of blasting vibration is an effective method to reduce the risk factor of blasting, optimize the blasting scheme and evaluate the safety level of blasting, and it is particularly important to study efficient and...

AI Technical Summary

Problems solved by technology

However, with the rapid expansion of the data volume of engineering blasting, the traditional stand-alone method of processing data sets will have serious problems, such as insufficient memory, long training time, etc.

Images