The ternary array passive positioning technique performs the estimation of target distance and bearing by measuring the time delay difference between array elements. The estimates are limited by the geometric model and time delay measurement errors, and the positioning accuracy and reliability are poor. In this paper, we propose an orientation matching-based motion target localization model. By introducing the target motion model, and establish a matching objective function for the time delay difference obtained by calculating the predefined target motion trajectory and the actual received data of the ternary array separately. The matching objective function is optimally solved by using Burg algorithm in the AR spectrum estimation method, and a highly accurate and reliable target parameter estimation can be obtained in a short time. The results of computer simulation experiments verify the effectiveness of the method.