First building the filter algorithm model of passive location by IRSTS by means of extended kalman filter, then aiming at the speciality of transcendental noise statistics and linearization error of measurement model effecting on filter precision during the study of extended kalman filter, adaptive extended kalman filter algorithm for passive location by IRSTS by means of subjuctive noise technique is advanced. It improved on extended kalman filter algorithm and approximated subjuctive noise statistics. The algorithm degraded linearization error and enhanced the nonlinear filter precision. The simulation experimental results show the advantage of adaptive extended kaiman filter algorithm under the same condition, the algorithm supplied practical value of engineering.