Research On Autonomous Exploration And Mapping Algorithm In Indoor Complex Scenes

With the continuous development of robot technology,service robots are more and more widely used in people's daily lives.The demand for service robots such as cleaning robots,elderly service robots,catering robots,etc.is increasing.Most of these robots work in complex indoor environments.It is necessary to build a map model of indoor scenes to complete the task.Therefore,it is a popular research in robotics to build a complete environment map model direction.Robot autonomous exploration mapping technology refers to the ability to autonomously generate composition points of the environment without relying on human control when the robot is in an unknown environment,to explore all areas of the unknown environment through continuous exploration of composition points,and to construct a complete map model of the environment.At present,there have been some developments in the independent exploration and construction technology,but there are still some problems.For example,the composition points generated by the robot autonomous exploration mapping technology cannot accommodate the robot,are prone to collision,and the composition points are inaccessible.The selection of the composition points is not optimal.These problems have caused the robot to automatically explore the map.Long time and low efficiency.This topic proposes a robotic autonomous boundary exploration mapping method based on the fusion of curve fitting,sliding window neighborhood planning and boundary fuzzy comprehensive evaluation,which optimizes the selection of composition points for unknown environments,guarantees the safety of composition points,and improves the robot The efficiency of independent mapping.The main research contents of this article are as follows:1)Investigate the current status of research in relevant fields at home and abroad,introduce the important algorithms and related technical theories of robot autonomous exploration and mapping,and introduce the hardware and software platform of robot autonomous exploration and mapping system.2)Propose a method based on curve fitting and sliding window neighborhood planning.Based on the Frontier-based algorithm,this method establishes a curve-fitting function onthe map boundary whose length is greater than the threshold,calculates the width of the boundary,and filters out unsafe composition points;and for the case where the target exploration point is in an unknown area and cannot be accessed,pass The sliding window neighborhood planning obtains the targets that can be reached in the free area to explore alternative points.Finally,the method is compared with the original method and other traditional methods in different scenarios,and the method has higher efficiency of autonomous mapping.3)Propose an autonomous boundary exploration method based on the fusion of fuzzy comprehensive evaluation method with curve fitting and sliding window neighborhood planning technology.First,establish the fuzzy membership of the map information gain,navigation cost,boundary length and neighborhood characteristics of the boundary in the safety boundary group Degree function,and then fuzzy comprehensive evaluation of the boundary by establishing a comprehensive evaluation matrix of boundary,mathematical operator model and normalization method,and finally the boundary with the best evaluation value as the target exploration boundary.Finally,a comparative analysis and verification experiment was carried out in actual scenes.This method explores fewer composition points to complete autonomous exploration and construction.