| dc.description.abstract | Railway curve realignment is critical for rectifying railway alignment deviations caused by excessive train load and repeated repairs. The existing realignment methods have limitations, such as low efficiency and precision, when considering realigning curves without key parameter information (CWI). To address the CWI issues, this study proposes a range identification and adaptive simplified particle swarm optimization (RI-ASPSO) algorithm combined with the existing principle of realigning railway curves. In this algorithm, the RI is designed to identify the range of curve parameters and is the premise of the ASPSO. Moreover, an automatic update strategy of the velocity threshold and an adaptive local random search strategy are developed in the ASPSO to efficiently and stably search the final near-optimal solution. The method is applied in real-world case studies, and the results show that the RI-ASPSO outperforms the particle swarm optimization (PSO) algorithm and coordinate method with higher accuracy, higher efficiency, and less deviation. | |
