Enhancing convergence speed and accuracy of virtual field optimization method for microseismic source location in tunnels

Abstract

Accurate and rapid microseismic source location is the foundation for tunnel rockburst warning. Here, we present a novel location strategy that employs an improved mayfly algorithm (IMA) to enhance the convergence speed and accuracy of the virtual field optimization method (VFOM) for tunnel microseismic events. By optimizing the initial position of the mayfly population and the moving velocity of mayfly personals, we develop an IMA with superior convergence speed in searching for source locations. The proposed method utilizes paired microseismic receivers in the tunnel monitoring array to create hyperbolic surfaces. Then, the IMA is employed to rapidly and accurately determine the intersection point of all hyperbolic surfaces, defining it as the microseismic source location. We compare the IMA-VFOM with different error levels in seismic wave velocity or arrival time against other traditional location approaches based on travel time differences. The results confirm that the IMA-VFOM’s convergence speed is, on average, more than 4 times that of the MA-VFOM algorithm. Compared to conventional methods, the IMA-VFOM method demonstrates higher location accuracy and stability. The average location error of the proposed method is $4.0043m$ when applied to real rockburst microseismic events in tunnels.