Abstract: To further explore the influence patterns of side anomalies in resistivity-based advance detection for roadways and to improve the detection accuracy,this study conducted three-dimensional forward modeling using an adaptive unstructured finite element method based on dual weighted posteriori error estimation.First,this study verified the applicability and accuracy of this method in the simulation of advanced detection using an infinite vertical plate model.Then,it established a model without considering side anomalies to simulate the advance detection response in roadways.Then,it conducted numerical simulations of high- and low-resistivity side anomalies' directions,the distances from side anomalies to roadways and mining faces,roadway sizes,power supply and reception ways,and the influence of side anomalies on advance detection in multi-layer media.Finally,this study analyzed the morphological characteristics of apparent resistivity anomaly curves.The results are as follows:(1)The actual positions of anomalies in front of roadways can be inferred from the extreme values of anomaly curves;(2)The responses of high- and low-resistivity side anomalies generally do not mask the response characteristics of anomalies in front of roadways.However,they can distort the anomaly curves in the corresponding range by making them convex or concave,with the distortion positions consistent with the actual positions of anomalies;(3)The presence of roadway cavities enhances the responses of anomalies on roadway floors.Although different roadway sizes impose different effects on anomalies on roadway rooves,they do not affect the responses of anomalies in front of roadways in general;(4)In multi-layer strata,anomaly response curves can still reflect the actual positions of side anomalies.Whether the response characteristics of anomalies in front of roadways are masked depends on the resistivity of each layer;(5)The identification of interference anomalies requires further research in combination with actual conditions.