Design Of Missing Radioactive Sources Searching Robot And Research On Sources Searching Algorithm

In recent years, nuclear industry developed rapidly. Meanwhile, radioactive sources lost from time to time. At present, people search for lost radioactive sources by themselves. It is inefficiency to search in this way and it is harmful for people’s health even lives. Nevertheless, if robots are used to search for those sources, it can be efficient during the processes and people can be protected. Therefore, designing a radioactive sources searching robot has great practical significance. Around the source searching task, this paper designed the structure of the radioactive sources searching robot, analyzed on kinematics by D-H method for the mechanical arm, put forward the source searching algorithm based on Gaussian Processes Regression, and confirmed the effectiveness of the algorithm by computer simulations and physical simulation experiments.Firstly, the robot designing index is put forward based on the analysis of present domestic and overseas radiation detecting robots and according to the sources searching task. In view of the designing index, the structure of the mobile platform is selected from common traveling mechanism. The mechanical properties of the double crawler walking mechanism is analyzed. Then, the structure of the mobile platform is designed.Secondly, the degrees of the mechanical arm which is installed on the mobile platform is determined by analyzing the radioactive sources grabbing task. The mechanical arm is decided to be RRR type from the common mechanical arm structure. Then, the mechanical arm is designed. The kinematics model of the arm is established and positive and inverse kinematics analysis are carried out. The size of the arm is determined according to the analysis of its working space.Next, radioactive sources searching algorithm based on Gaussian Processes Regression is proposed. The algorithm is been divided into two parts. The first part is environment radiation intensity distribution estimating algorithm. Gaussian Processes Regression is used to estimate the radiation intensity of unknown points when given the coordinates and radiation intensity of known points as training set. The second part is path planning algorithm of the sources searching robot. An objective function based on Gaussian Processes Regression of an optimization problem is established. The robot finds the optimal target position in the adjacent part by solving an optimization problem. As a result, the globally optimal solution is obtained by repeatedly local optimization.Finally, because radioactive sources are extremely dangerous, two simulation experiments are taken to verify the effectiveness of the proposed algorithm. In the computer simulation, a virtual gamma radiation field is build. 20 experiments are conducted in this field and the quality of the estimation is analyzed by the simulation results. In the physical simulation experiments, the gamma radiation is simulated by an electromagnetic radiation of 2.4 GHz. The results of two outdoor experiments verify the effectiveness of the proposed algorithm.In general, the structure of a radioactive sources searching robot is designed in this paper, and the kinematics analysis of the mechanical arm is completed. Meanwhile, an effective radioactive sources searching algorithm is proposed. Theoretical foundation is established for searching radioactive sources by autonomous robots.