The classical Kalman filtering algorithm requires the system to be linear, but in reality, the system is often nonlinear.Therefore, if the nonlinear system is expanded by Using Taylor series and then the first term is taken, then the nonlinear system becomes approximately a linear system. In this way, the system state prediction information can be deduced recursively according to the Kalman filtering framework, which is the EXTENDED Kalman filtering algorithm based SLAM (EKF-SLAM).Extended Kalman filter is a suboptimal state prediction method based on Bayesian theory.