Construction Of TBM Operation Parameter Decision Optimization Model Based On Genetic Algorithm

The tunnel boring machine(TBM)is an underground tunneling engineering equipment in which the motor and hydraulic motor propel the cutter head,the hob squeezes the rock,and the support and slag discharge are carried out at the same time.During the TBM driving process,the setting of the TBM operating parameters is still in the manual control stage,and the unreasonable setting of the parameters by the operator causes the TBM to advance too slowly in a non-optimal driving state.Based on the purpose of accelerating project progress,reducing project investment costs,and increasing intelligence.The optimization decision-making problem of TBM operating parameters has become one of the hot topics in the intelligent tunneling research of tunnel engineering.Based on the actual measured data of water supply project,this paper integrates gradient lifting decision tree,random forest,genetic algorithm and other algorithms to build a TBM operating parameter optimization decision model with TBM advancement speed as the optimization goal to assist the operator in decision-making and establish TBM intelligent tunneling system.The main research work of this article is as follows:1)Based on the graph analysis method,the relationship between TBM tunneling parameters,surrounding rock grades,and TBM propulsion speed was studied.It is concluded that there is a complex nonlinear relationship between the surrounding rock grade,TBM advancement speed and TBM driving parameters,which indicates the direction for the subsequent model establishment.2)Based on the state discriminant function,TBM operating data is extracted from the original data,and data preprocessing is performed on the operating data.The gradient lifting tree algorithm is used to construct the surrounding rock grade prediction model,and the surrounding rock grade prediction accuracy reaches 99.8%.A neural network and random forest algorithm are integrated to construct a TBM propulsion speed prediction model.The average absolute error of TBM propulsion speed prediction is 10.2%.Based on the above model,the results of the surrounding rock grade prediction model are used as input parameters,the TBM propulsion speed prediction model result is the optimization goal,and the genetic algorithm is the intelligent search algorithm.The TBM operating parameter optimization decision model with the propulsion speed as the optimization goal is constructed.The model can output the optimal operating parameter combination within 1.5s.3)Based on the LSTM neural network algorithm,the TBM posture prediction model is constructed.The model takes the data before the real-time tunneling point as input and predicts the posture change after the tunneling point,so that the operator can predict the posture change in advance and take relevant measures to avoid the occurrence of serpentine tunnel.