<p>Global Navigation Satellite Systems (GNSS) are essential for transportation, communication, and scientific research, yet their accuracy and reliability are frequently compromised by the effects of space weather. Disruptions to the ionosphere caused by space weather affect GNSS signal propagation paths, leading to delays, attenuation, or even interruptions, which significantly degrade navigation accuracy and compromise service reliability. Conventional models often fail to accommodate the complex, nonlinear ionospheric variations driven by these disturbances. To overcome these limitations, the Lagged Response-Enhanced Y-Structure Transformer (LRYformer) is introduced, offering a novel framework that integrates multiscale processing and accounts for the delayed influences of solar and geomagnetic activity on GNSS performance. By employing a lagged response mechanism, Y-structured attention-based feature extraction, and dimensionality reduction of historical data, LRYformer demonstrates exceptional predictive capability under diverse and extreme space weather conditions. Experimental evaluations reveal its superiority over existing models in terms of accuracy and robustness, particularly during challenging events. Furthermore, the model effectively addresses the temporal lags inherent in ionospheric responses, reducing delays and improving predictive efficiency for complex scenarios. The development of LRYformer represents a significant methodological advancement for analyzing the impacts of space weather on GNSS systems. Its ability to enhance predictive precision and system reliability underpins its potential to strengthen satellite navigation performance across a wide range of critical applications.</p>