Research On SLAM Of Indoor Robot Based On Lidar

Simultaneous localization and mapping,also known as SLAM which is a key prerequisite of truly autonomous robots.With SLAM technology,mobile robots can have the ability to determine their location and build a surrounding environment map in an unknown environment without relying on an external reference system.This paper is based on the SLAM problem of indoor mobile robots,the main research is as follows:Firstly,expounds the current development status of autonomous mobile robot technology and SLAM technology,establishes the kinematics model of two-wheel independent drive robot,and studies the track estimation algorithm and odometer calibration algorithm.Secondly,the research of lidar ranging algorithm is carried out,mainly completed the work of designing the motion distortion removal algorithm and a quadratic interpolation improvement algorithm is proposed,which can effectively overcome the influence of motion distortion on measurement results and further improve measurement accuracy.Thirdly,the SLAM algorithm based on RBPF theory is studied.According to the characteristics of indoor mobile robots,the problem of particle dissipation in RBPF-SLAM is improved.It takes into account the observation and odometer information of lidar.Firstly,the sampling information of lidar is used to limit the sampling interval to a smaller one scope,followed by weight statistics and evaluation of particles collected in the interval,which delays particle dissipation and reduces the number of particles required,improving computational efficiency.Aiming at the improvement of traditional data fusion algorithms,a more effective method is proposed.The odometer and scan-matching information is integrated,and the traditional odometer model is used to sample the particle to obtain the proposed distribution when the scan-matching fails,which improves the accuracy and robustness.Fourthly,the map construction algorithm is studied.The overlay grid construction algorithm is used to realize the construction of the environment map,and the simulation experiment is carried out through the ROS system.The experimental results show that with the same number of particles,the improved algorithm obtains a more accurate environment map and exhibits better robustness.