Abstract: In order to solve the problem of driver comfort reduction or even dangerous situations due to late recognition of the cut-in of side-lane vehicles of the adaptive cruise control system, this paper proposes a control strategy based on a support vector machine (SVM) to recognize cut-in vehicles and adjust the weight parameters of the model predictive control (MPC) in real time. Firstly, lane change characteristics of real traffic vehicles are extracted, and the SVM algorithm is used to train the recognition model of the cut-in front vehicle. Then, the weight adjustment mechanism is designed based on the fuzzy control theory, and the MPC controller is optimized to obtain the expected acceleration. Finally, a hardware-in-the-loop test platform is set up in Prescan, Matlab/Simulink and NI real-time system environments, and different cut-in working conditions are compared and analyzed. The results show that the adaptive cruise control strategy can switch the following target in advance, reduce the fluctuation of longitudinal acceleration and avoid forced vehicle braking, which improves the comfort and safety of the system.