Intelligent velocity picking based on Simulated Annealing Particle Swarm Optimization algorithm

In seismic exploration, velocity information is crucial for deciphering the subsurface geology. The traditional method is to pick up after calculating the velocity spectrum, which puts forward higher requirements for the accuracy of the velocity spectrum. This paper proposes an innovative method for determining stratigraphic velocities without computing the velocity spectrum. The method begins with the initialization of picking point coordinates, leveraging the free-search capabilities of the Simulated Annealing Particle Swarm Optimization (SAPSO) algorithm to facilitate information exchange among the points and to avoid entrapment in local optima. Then, the algorithm's global optimization strategy is refined to maintain the individual movement trends of each point, accommodating the multi-objective attribute of the picking process. Finally, the fitness criteria and the objective function's form are defined during the search to evaluate the picking effect. Both synthetic and field data examples demonstrate that, compared with two velocity spectrum clustering picking algorithms and the dynamic time warping algorithm, the proposed method requires no velocity spectrum calculation while achieving a 0.4% reduction in the mean relative error of the picking velocity and anisotropy parameter.