Abstract: In order to reduce the influence of sensor systematic errors, a mathematical model of odometry systematic error and laser rangefinder installation error of differential-drive mobile robot is established firstly. Then, an iterative calibration method based on extended Kalman filter algorithm is proposed for odometry systematic error calibration and laser rangefinder installation error calibration, which can calibrate these two sets of errors during localization in real time. When verifying the method by simulation, the errors estimation can effectively converge to the true values of the errors. While in physical experiments, the errors estimation can effectively converge, and the error of dead-reckoning is greatly reduced after calibration.