Research On Source Location Method Of Mobile Robot Based On Contour Gradient

In daily life,there are many hidden dangers of leakage of hazard sources,which always pose a threat to the safety of life and property of national and personal.In order to quickly and accurately determine the location of hazard sources,so as to reduce the harm brought by them,it is necessary to conduct in-depth research on the positioning method of hazard sources.Although the traditional source localization method has some disadvantages,such as low efficiency and poor flexibility,these disadvantages can be effectively avoided by using mobile robots.Therefore,this topic uses a mobile robot equipped with photosensitive sensors as the experimental platform.Aiming at the path planning problem brought by the actual search area,the source location method of the robot in the unknown scene is studied,which provides convenience for the processing of the danger source in the later stage.The main research work of this paper is as follows:(1)The mathematical model of the mobile robot platform was established,and the software framework of the source localization method was built.STM32 is used to build the light source data acquisition module,and the data acquisition is realized.In the process of robot motion,The robot's position data is combined with the light intensity data,the two kinds of data matching is realized.(2)In order to obtain the radiation intensity contour distribution of the source in the region by using the sampling data of the robot,this paper studies the contour drawing based on Kriging interpolation prediction algorithm.This algorithm can use the limited sampling data to predict the data of each grid cell within the target region,so as to obtain the contour distribution of the source within the region.In order to improve the efficiency of the robot in the data sampling stage,two kinds of traversal methods,square and diagonal,were used to carry out experiments.Experimental results show that the proposed method can obtain contour distribution effectively.(3)Based on the contour distribution predicted by robot,this paper proposes two source positioning methods based on Kriging interpolation prediction for source positioning.An ergodic source localization method is proposed for multi-source and single-source localization in a small area.An adaptive gradient extremum source localization method is proposed to solve the problem of single source localization in a large area.Experimental results show that the proposed method can successfully locate the source.(4)An improved hybrid path planning algorithm for robots,RRT algorithm after path smoothing and improved DWA algorithm,are proposed to solve problems of robot's exploration of unknown environment and obstacle avoidance in environment with temporary obstacles.Since the traditional RRT algorithm has the disadvantage of high path cost,smoothing the tree nodes on the path can effectively shorten the path length.In order to make all feasible paths in the DWA window take the same proportion in the evaluation function,the three evaluation functions of the algorithm are normalized,so that the robot can pass through the obstacle area with a better path.To sum up,the robot traversal strategy was used to collect data in this project,and the contour distribution was obtained through Kriging algorithm.The source location method uses the predicted contour distribution to determine the target points for the path planning algorithm,thus guiding the robot to complete the source location task.Finally,the mobile robot platform was used for experimental verification in the laboratory.