To quantitatively analyze the interaction between vehicles in traffic flow, the vehicle interaction velocity-changing model of arterial road was studied. The concept of potential field influence region was defined through the analogy of gravitation and repulsion in the potential field theory. The attraction and repulsion between the target vehicle and the surrounding vehicles were attributed to the changes of the interaction area affected by the potential field. A vehicle interaction analysis method considering the influence area of potential field was proposed, and a linear model of the interaction area between target vehicle velocity variation and potential field was established. Vehicle coordinates and speed data collected by the P3-DT BeiDou high-precision positioning direction finder were used to calibrate model parameters. The speed changes of the target vehicle were calculated by the proposed method and compared with the measured data. Results show that the error values between the calculation results and the measured data were less than 15%. When the lane-changing time became shorter, the interaction between the lane-changing vehicle and the rear vehicle on the target lane was stronger, and the deceleration of the rear vehicle was operated more quickly, which verifies the validity of the model. The model normalizes the vehicle speed and spacing in traditional microscopic traffic flow analysis into the interaction area of potential field, which can provide method for the study of multi-vehicle interactions in microscopic traffic flow and speed control strategy for automatic driving vehicles.