Addressing the issues of inadequate positioning accuracy and limited real-time performance in traditional gas leakage source detection, this study introduces an arrival time-difference localization method that integrates the enhanced sparrow algorithm. The algorithm addresses the issues of poor noise immunity and inadequate accuracy in the arrival time-difference localization method by integrating the sparrow algorithm. Additionally, homogenized logistic mapping is incorporated to enhance the diversity of the population. Moreover, a nonlinear convergence factor is introduced to accelerate the iterative process. This enhancement boosts the global search capability in the initial stages of the algorithm, enhances convergence accuracy in the later stages, and alters the traditional spar-row algorithm, which tends to get stuck in local optimal search scenarios. Through simulation verification, the enhanced algorithm outperforms the traditional algorithm with a noise index lower than 0.15. The error rate is reduced by approximately 14%. When the noise index reaches 0.15, the error rate decreases by about 21%. The enhanced algorithm demonstrates superior performance and increased robustness.