基于传感器布设及粒子群算法的破裂源定位研究

Research on Rupture Source Localization based on Sensor Arrangement and Particle Swarm Algorithm

岩体破裂源的精确定位是岩体破裂场分析、动力灾害预警的基础,而基于分区波速模型的传统定位方法在倾斜层状岩体中会产生较大的误差。因此,本文详细分析了传统定位方法中传感器布置方式对定位精度的影响,发现传统定位方法的定位精度严重依赖于传感器的排布方式。通过自适应变异及线性动态惯性权重对粒子群算法进行优化,并结合最小走时的计算方法可将定位平均误差控制在2 m以内,定位精度显著提高,且优化后的粒子群算法受传感器排布方式的影响较小,具有良好的稳定性。

Precise localization of rock rupture sources is the basis of rock rupture field analysis and dynamic hazard warning,while the traditional localization method based on sectional wave velocity model will produce large errors in inclined layered rock masses.Therefore,the influence of sensor arrangement on localization accuracy in the traditional localization method is analyzed in detail,and it is found that the localization accuracy of the traditional localization method depends heavily on the sensor arrangement.The particle swarm algorithm is optimized by adaptive variation and linear dynamic inertia weights,and the average positioning error is controlled within 2 m by combining the calculation method of minimum travel time,which significantly improves the positioning accuracy,and the optimized particle swarm algorithm is less affected by the sensor arrangement,and has good stability.