Abstract: Intelligent vehicles have poor tracking effect under complex traffic conditions.To address the problem，this paper proposes a discontinuous curvature trajectory tracking control strategy considering vehicle steering safety.Based on the influence of road curvature and vehicle speed index，an improved pre-aiming error model is designed.Combined with the safety speed constraint，an adaptive presight controller is built.The pre-sight path and road curvature are obtained by five-order polynomial fitting.According to the pre-sighted position geometry，the driver steering model is designed and the front wheel angle equation is derived.Matlab／Simulink，Carsim and Prescan are employed to build a joint simulation platform，and the double line shift，discontinuous curvature trajectory and continuous lane change conditions are selected for comparative experiments.Our results show under the above working conditions，the control algorithm achieves the adaptive adjustment of the pre-aiming path with the change of external conditions，and it improves the tracking accuracy by about55.34%，65.86% and 46% respectively while ensuring the stability，safety and adaptability of the vehicle.