Multisensor Fusion SLAM Research Based on Improved RBPF-SLAM Algorithm
Algorithm 1
The specific flow of the EKF-based multisensor fusion algorithm.
EKF-based sensor fusion algorithm for wheeled odometer and IMU
Step 1 Acquire sensor data from the wheeled odometer and IMU.
Step 2 Construct an EKF using a nonlinear model of a wheeled odometer.
Step 3 Start status updates to the system and add system noise.
Step 4 Update the system state quantities and the system covariance matrix by combining the state quantities of the previous moment and listening to the odometer information as the observed quantities and the observed covariance matrix.
Step 5 Update the system state quantities and system covariance matrices obtained in Step 4 with the state by listening to IMU information as observation quantities and observation covariance matrices.
Step 6 Use the fused system state quantities and covariance matrix as the initial bit poses of the SLAM algorithm.
