Research On Autonomous Navigation Method In Complex Environments Of Indoor Robot

With the boom of artificial intelligence and the development of robotics,the core technology of autonomous navigation has become the focus of research in the field of indoor robotics.Because the indoor environment is relatively complicated,and there are many obstacles and random distribution,it is necessary to face many practical problems to achieve autonomous navigation in a complex indoor environment requires.Therefore,it is of great engineering practical significance and practical application value to study the indoor robot autonomous navigation system in complex environment.The main work of this thesis is as follows:The SLAM algorithm based on improved Rao-Blackwillised particle filter RBPF is studied.Aiming at the low precision and severe particle dissipation of the traditional RBPF-SLAM algorithm,this paper studys an improved RBPF-SLAM algorithm,which uses the Point-to-Line Iterative Closest Point(PL-ICP)to register the laser scanning data of two adjacent frames,and the inter-frame matching result replaces the odometer reading to optimize the proposed distribution,and a particle weight balance strategy is introduced in the resampling process to alleviate the particle dissipation.Through simulation and comparison experiments,the improved algorithm improves the computational efficiency while maintaining the diversity of particles and obtains more accurate robot positioning and higher precision maps.The path planning algorithm combined with improved A* and dynamic window approach(DWA)is studied.Aiming at the limitation that the traditional A* algorithm could not be applied in the complex indoor environment,this paper studys a global planning algorithm for improved A*,The simulation experiment shows that the improved algorithm reduces the computational complexity and improves the search efficiency.The experimental results show that the algorithm is effective.According to the characteristics of many obstacles in the complex indoor environment,this thesis adopts the local planning algorithm of DWA,and selects the evaluation function suitable for this topic for the DWA algorithm through simulation experiments.Combining global planning with local planning,it could make use of known maps and meet real-time requirements to achieve better planning results.The indoor robot autonomous navigation system is designed and implemented.Firstly,the robot operating system(ROS)is used as the software platform,the two-wheel differential robot is the hardware platform,and the autonomous navigation system is designed with the laser radar as the main environmental information acquisition device and the wheel odometer as the auxiliary equipment.Then based on the SLAM algorithm and path planning algorithm studied in this paper,SLAM contrast experiment and autonomous navigation contrast experiment in three different levels of indoor environment realize the mapping of indoor unknown environment and autonomous navigation and obstacle avoidance.The results show that the studied algorithm and the designed navigation system have high efficiency and good robustness,and could complete the task of dynamic obstacle avoidance during navigation,which has certain practical application value.