Research On Radiation Map Construction Technology Based On Multi-source Information Perception

When exploring unknown radiation environments,determining how special environment robots can accurately establish environment maps,collect radiation information,and transmit processed radiation data to the navigation system is a key challenge and focus of research on special environment robots.Therefore,this study focuses on research into mobile robot systems,autonomous localization and navigation technology,and radiation field map construction in order to solve issues such as insufficient positioning accuracy of mobile robots,construction of radiation field maps,and inability to avoid dangerous areas in complex radiation environments.By improving the environmental perception performance and local autonomy of operating robots,we aim to enhance the efficiency and safety of operations and play an important role in practical work.The main research work of this subject is as follows:(1)Analyze the robotic system and establish a kinematic model to provide theoretical motion calculation basis for the odometer estimation and autonomous navigation of the robotic system;establish relevant sensor error models to reduce the data errors read by the robotic system,and analyze the sensor calibration theory to provide the relative position relationship of each sensor for the SLAM algorithm;based on the ROS architecture,according to the communication protocol of each sensor,construct a robotic software system framework with driving,perception,and decision-making capabilities.(2)In non-structured environments,a single sensor usually cannot obtain sufficient information to accurately locate and map.Therefore,in response to the problem of inaccurate positioning and mapping of a single sensor,this project combines data fusion theory to construct a tightly coupled odometer fusion algorithm,and combines it with the SLAM algorithm,and the resulting robot positioning accuracy and the constructed environment map can meet the needs of practical projects.(3)For the key technology of radiation field map construction,this project studies radiation field prediction methods and builds a neural network-based prediction model to predict radiation information in the detected environment.Then,a decision-making algorithm is designed to filter the radiation information and send it to the navigation map.Finally,combined with the environmental geometric map information and radiation information,a radiation map distribution is constructed,which can be applied to the navigation system of the robot to automatically avoid dangerous areas during task execution.(4)Experimental results and analysis.To verify the radiation map construction capability of the robotic system and its ability to autonomously avoid dangerous areas,relevant experiments are conducted in corridors,corners,and outdoor environments.Simulation experiments are carried out based on the Gazebo and RVIZ tools under the ROS framework,and an experiment was conducted to construct a radiation field map using the mobile robot system developed in this study.The results showed that the developed tightly coupled odometry algorithm improved the accuracy by 65.74% when the robot’s speed was limited to 2 m/s;the average percentage error of the designed neural network model is 0.938%;the position accuracy of the predicted radiation field map is 0.393m;The robot system framework,SLAM algorithm,and radiation field mapping technology studied in this research can accurately establish radiation field maps.In addition,a set of robot systems developed in this study has the ability to explore unknown environments autonomously and avoid dangerous areas.