The rapid process of urbanization and industrial development has raised significant concerns regarding the presence and management of hazardous substances. However, conventional methods employed for identifying hazardous substances and monitoring urban safety often suffer from low efficiency and accuracy. This paper proposes a novel approach that combines deep learning and genetic algorithms, which utilizes the Bidirectional Long Short-Term Memory model to capture temporal features in hazardous substance data and introduces the Attention Mechanism for weighted processing of crucial information, thereby improving recognition capability. Genetic Algorithms are employed to optimize the performance and generalization capacity of the deep learning model. Experimental validation demonstrates that the proposed approach achieves higher accuracy and faster processing speed, effectively enhancing urban safety monitoring. This research holds practical implications for urban safety management and accident prevention, offering an innovative solution to guarantee urban safety.

Complex fluid–particle interactions are ubiquitous in natural environments and engineering applications, with their underlying mechanisms often attributed to interparticle attraction and repulsion. To understand the interaction mechanism between the dual particles, this study examines the setting process of dual particles using the Immersed Boundary-Lattice Boltzmann Method (IB-LBM), with a focus on the effect of the density difference between particles. Two typical configurations—tandem and side-by-side—are considered in the analysis. In the tandem configuration, when ${\rho }_{LP}/{\rho }_{TP}<1$, the TP inevitably kisses the LP due to its greater settling velocity, thus initiating the classical drafting-kissing-tumbling phenomenon. As the density of the TP further increases, the attractive effect exerted by the LP on the TP becomes weak. Conversely, when ${\rho }_{LP}/{\rho }_{TP}>1$, kissing between two particles is mainly determined by the density of LP. Whether kissing occurs between the two particles depends on a critical value ${\rho }_{LP}/{\rho }_{TP}=1.2:1.14$. Although the LP’s attraction to the TP strengthens with increasing LP density, beyond this certain threshold, this attraction becomes insufficient for the TP to catch up with the LP. In a side-by-side configuration with two particles of different densities, their interaction evolves from initial attraction to subsequent repulsion. This phenomenon is not observed in pairs of particles with identical density. Moreover, with increasing density difference between the particles, the attractive effect from the higher-density particle on the lower-density one strengthens, whereas the repulsive interaction between them gradually weakens. When the particle density ratio reaches $1.4:1.14$, the lateral migration of the particles becomes very small; although they still interact with each other, the effect becomes extremely weak. This work systematically elucidates the influence of density disparity on particle interaction, providing insights into understanding more complex multiparticle system dynamics.