The application of extended kalman filtering based on SLAM

Abstract

EKF (Extended Kalman Filter) is a non-linear state estimation algorithm based on the development of Kalman filtering technology. In SLAM (Simultaneous Localization and Mapping), EKF technology is widely used to realize robots' independent positioning and map construction. EKF technology mainly estimates the state of the robot by using the information provided by the sensor, and then realizes the robot's self-positioning and the environment's map. This paper uses EKF algorithm for testing, and analyses the simulation results, which compared with Kalman filtering technology, the main difference between the two is that EKF can handle non -linear dynamic systems. In SLAM, robots often detect environmental detection through sensors such as laser radar, camera, and then update the position and posture of the robot by the receiving sensor data. In this process, the EKF technology can perform non -linearity by sensor data by sensor data Transformation makes the state of the robot more accurately estimate. Specifically, Extended Kalman Filtering technology can be divided into two steps: prediction and update. In the predicted phase, EKF uses the current robotic status and motion equation to predict the robot status of the next moment; in the update stage, EKF uses sensors to measure data to update the current state of the robot. By repeating the prediction and updating two steps, you can get the trajectory of the robot movement and the map of the environment where the robot is located.