A machine jam risk early warning method and system for a tunnel boring machine

Abstract

The application provides a tunnel boring machine jam risk early warning method and system, the method comprises the following steps: obtaining geological section parameters based on the three-dimensional geological model of the current and the set distance in front, and simulating the interaction between the shield and the surrounding rock by the finite element analysis method to obtain the predicted segment theoretical advancing resistance and cutter head torque; calculating the cutter head rotating resistance, actual advancing resistance, effective rock breaking torque and effective rock breaking thrust based on the historical and current tunneling data to construct a time series set, and using a sliding window linear regression to calculate the change trend slope of the rock breaking resistance and thrust; inputting the above-mentioned geology, simulation, real-time mechanics and dynamic trend characteristics into a pre-trained gradient boosting decision tree model to output the jam risk level; comparing the estimated shield body and cutter head failure time based on the preset rules and threshold; and comprehensively generating and displaying the graded early warning results. The application realizes early graded early warning of the full-face tunnel boring machine jam risk, and overcomes the disadvantages of the traditional method which relies on static threshold and artificial experience.

Images