Research Of Plume Discovery Strategy Of Cross Searching Based On A*-Ant Colony Algorithm

Plume discovery is the initial stage of plume tracking and leakage source location,which determines the basic time of the subsequent search process.This subject researches the problem of plume discovery in a windless environment,and the main factors that affect the real-time and stability of plume discovery in the process of plume discovery:the length of the robot’s traversal path,the search strategy of traversing different target points,and the problem of accurate target tracking After analysis,a cross-search plume discovery strategy based on A*-ant colony algorithm is proposed,which shortens the longest discovery time and improves the stability of the plume discovery search time.The specific research content of this article on the discovery of plumes is as follows:(1)Proposed A*-ant colony algorithm path planning.Analyze the shortcomings of traditional A*algorithm and ant colony algorithm in path planning,such as the search angle being affected by the thickness of the grid map grid division,combine A*algorithm and ant colony algorithm to multi-target path planning,and smooth the path Eliminate the influence of search angle and generate the shortest path between multiple target points.This method is applied to the path planning of the plume discovery process to shorten the plume discovery time and improve the real-time performance of the plume discovery.(2)A cross search strategy is proposed.The status quo of domestic and foreign research on plume discovery is analyzed.Through the research and analysis of commonly used plume discovery algorithms,the difficulty and key points of the plume discovery task are clarified.According to the blindness of the smoke plume discovery search,the known environmental information is used to create a risk coefficient model of suspected leak sources,so that the sequence of traversing different suspected leak sources is more suitable for the actual situation.A cross-search strategy is proposed.By establishing primary and secondary monitoring points for suspected leaking sources,a cross-search of rough search and precise search for suspected leaking sources is realized.The rough search avoids the worst case,and the precise search makes A certain probability is detected when the suspected source gas leaks,which reduces the longest discovery time during the plume discovery process.(3)The problem of feature extraction and matching in target tracking is studied.In the process of plume discovery,the robot needs to perform image matching through vision,so as to analyze whether the search target is lost and judge whether the robot deviates from the path trajectory.This paper analyzes the current classic image matching methods,and then proposes a fast image matching algorithm for sub-scale corresponding regions,so that the robot can accurately track the target in real time during the plume discovery process,prevent the robot’s trajectory from deviating from the planned path,and improve the plume Real-time search.(4)Use MATLAB software and turtlebot2 robot to simulate or experiment with the algorithms in each chapter,and compare them with traditional algorithms.Experimental results show that in simulation and actual environments,the cross-search plume discovery strategy algorithm based on A*-ant colony algorithm proposed in this paper has better stability,and the average stability is increased by 74.54%compared with traditional plume discovery algorithms.The average plume discovery search time is 35.65%lower than the traditional algorithm.The longest discovery time is 51.56%lower than the traditional algorithm.This paper studies the advantages and disadvantages of the smoke plume discovery algorithm,and studies the blindness,low search efficiency,and poor stability of the plume discovery search in practical application scenarios,and provides for the plume discovery based on the robot platform in the actual scene in the windless environment For reference.Experiments prove that the cross-search plume discovery strategy algorithm based on A*-ant colony algorithm proposed in this paper can effectively shorten the longest plume discovery time,improve the efficiency of robot early warning,avoid serious consequences,and have better stability.