Reinforced concrete (RC) slabs, as the core load-bearing components in construction engineering, are prone to internal damage induced by impact loads, and accurate positioning of impact locations is a key task in structural health monitoring. The proposed method was developed for typical RC slabs such as building floors, bridge decks, and road slabs. Traditional acoustic emission (AE) positioning methods suffer from low positioning accuracy and a tendency to fall into local optimum when applied to RC slabs, which is attributed to the material’s heterogeneity, the complex propagation characteristics of stress waves and ambient noise interference. In this study, a GA-PSO hybrid algorithm is proposed, which integrates the global search capability of the Genetic Algorithm (GA) with the superior local convergence performance of the Particle Swarm Optimization (PSO) algorithm. The premature convergence issue of the traditional PSO algorithm is alleviated by adopting strategies including tournament selection, α hybrid crossover, boundary-constrained mutation, and linearly decreasing inertia weight. Based on the Time Difference of Arrival (TDOA) principle, the root mean square error between the theoretical and measured time differences is taken as the fitness function, and a boundary penalty mechanism is incorporated to ensure the physical validity of positioning results. AE data were acquired through drop weight impact tests to verify the performance of the proposed algorithm. Compared with traditional TDOA grid search, pure GA, and pure PSO methods under the same conditions, the proposed GA-PSO algorithm achieves an average localization error of only 54.95 mm, which is 61.0% lower than that of pure GA, while reducing the error standard deviation from approximately 114 mm to 24.87 mm. The average positioning error for all impact sources on the RC slab is within 100 mm, with the error in the central area as low as 42.97 mm. These results demonstrate that the GA-PSO algorithm significantly outperforms existing methods in terms of accuracy, stability, and maximum error control, verifying its high potential for impact source localization in complex heterogeneous materials.

Crack coalescence between discontinuities plays a critical role in the progressive failure and stability evolution of jointed rock masses. This study aims to develop a modified coalescence coefficient for quantifying crack coalescence behavior and evaluating its influence on rock mass stability. Uniaxial compression tests and heterogeneous numerical simulations were conducted to investigate the effects of crack dip angle, rock bridge dip angle, crack length, and rock bridge length on crack coalescence. The results revealed cubic polynomial relationships between the coalescence coefficient and crack dip angle, rock bridge dip angle, and crack length, whereas a linear relationship was identified with rock bridge length. Based on an L₂₅(5⁴) orthogonal design and nonlinear regression analysis, a modified coalescence coefficient applicable to jointed rock masses under static loading conditions was proposed. The coefficient was subsequently incorporated into the block theory implemented in GeoSMA-3D and applied to a tunnel excavation case. The results demonstrated that considering structural-plane coalescence significantly alters the geometry and volume of key blocks, leading to a more realistic stability assessment. The proposed coefficient provides an effective tool for improving block identification and stability evaluation in underground rock engineering.