THE PREDICTION OF EPB-TBM PERFORMANCE USING FIREFLY ALGORITHMS AND PARTICLE SWARM OPTIMIZATION

Abstract

Penetration rate is one of the most important parameters in determining excavation time in tunnelling operations. Providing a prediction model or a mathematical relationship can give a better understanding of this issue. A mathematical equation between input and output parameters can be optimized by using algorithms such as Particle Swarm Optimization and Firefly. Since drilling operations interact between ground and machine, therefore, the effective parameters on the penetration rate are divided into two general categories such as machine and geological factors. Effective geological factors include internal friction angle, cohesion, specific gravity, shear modulus and groundwater level. In addition, the important parameters of TBM are torque, thrust jacks, and rotation speed. By defining an initial mathematical function, two optimization algorithms, which look for the most optimal mode, the goal here is the same as the mean square error (MSE). Finally, by examining and comparing the performance of two algorithms, using the coefficient of determination and the mean square error, it found that the Firefly algorithm has a better performance than the Particle Swarm Optimization algorithm.

Keywords

• EPB-TBM
• penetration rate
• regression
• firefly algorithm
• particle swarm optimization