Abstract: A lane-changing path planning algorithm based on the Frenet coordinate system is proposed to address the safety, comfort, and real-time issues of intelligent vehicles during lane changing in highway environments. Firstly, cubic spline curves are used to fit reference lines to lane-line information obtained from vehicle-mounted cameras. In the Frenet coordinate system, vehicle motion is decoupled longitudinal and laterally, and a fifth-degree polynomial lane changing candidate trajectory cluster is generated through time and position point sampling. Secondly, the cost function with acceleration change rate as the core is used to evaluate each trajectory, and considering maximum curvature, maximum acceleration, and collision constraints, the optimal lane-changing trajectory is selected. Finally, the feasibility of planning the trajectory is verified using a multi-point preview lateral control algorithm. The simulation results show that the lane changing path planning algorithm can generate safe, comfortable, and collision-free lane-changing trajectories in real-time, meeting the requirements of real-time lane changing.