Research On Mobile Robot Autonomous Exploration Mapping And Path Planning Technology

In the era of intelligence,mobile robots play a crucial role in various applications.To enable their intelligence,two key components are essential: autonomous mapping technology and path planning technology.Currently,manual mapping methods are used by companies to construct maps of unknown environments.However,these methods are time-consuming,laborintensive,and often lack accuracy.This thesis aims to address these limitations by introducing the edge detection algorithm based on boundary wavefront detection and proposing an exploration strategy that incorporates directional information and a topological map.This novel approach significantly reduces the time required for map construction while enhancing its accuracy.Once the map construction is complete,the path planning algorithm is executed on the generated map.Conventional A~* algorithms plan paths in proximity to obstacles,resulting in non-smooth trajectories that are difficult for robots to follow.On the other hand,the DWA algorithm tends to encounter collisions with obstacles during the steering process.To overcome these challenges,this thesis proposes improvements to the A~* and DWA algorithms,building on existing state-of-the-art techniques.These enhancements aim to enhance the planning and obstacle avoidance capabilities of mobile robots in complex environments.The main contributions of this thesis cover the following aspects:For the problem of inefficient and incomplete map construction of mobile robots in complex unknown environments,a boundary exploration algorithm based on wavefront edge detection is used.The boundary exploration strategy of this algorithm has local minima and only considers the distance and size of unknown space,which leads to low exploration efficiency in complex environments,and this thesis proposes an exploration strategy based on direction information and topological maps.When selecting the next boundary target point,not only the size of the target point’s information gain is considered,but also the direction of the mobile robot;the coordinate points of the mobile robot’s location and all boundary target points are connected to build a topological map,and the visited boundary target points are marked to reduce repeated exploration of the unknown area,and the improved exploration construction method can balance map coverage and map construction time to a greater extent.The feasibility of the algorithm is verified by experiments,and this method improves the efficiency and accuracy of map construction,and the constructed maps can be used for navigation tasks of mobile robots.When the mobile robot uses the A~* algorithm to plan the path,the generated path point is close to the obstacle and the path is not smooth.This thesis proposes the obstacle distance evaluation function and introduces this sub-function into the heuristic search function to avoid collision of the mobile robot because the path point is close to the obstacle.On this basis,the bounding box method is proposed to remove redundant nodes at the path point and keep the path planned by the mobile robot can avoid collision and facilitate trajectory tracking.Experimental results show that the improved algorithm solves the problem that the path point of the mobile robot is close to the obstacle and thus collision occurs,and smooths the path.The mobile robot may collide with obstacles in the process of moving and turning,so the DWA algorithm is optimized to add the cost of the mobile robot footprint to the original evaluation function of the algorithm to avoid collision when the mobile robot turns at a corner.For the problem that the DWA algorithm falls into the local optimum,this thesis combines the global path of the A~* algorithm,which makes the mobile robot treat the global path point as the local target point,and adds the path distance evaluation function to effectively reduce the probability of gyration situation.This thesis proposes a speed sampling space extraction strategy,and through the comparison experiments of DWA algorithm,improved DWA algorithm,MPC algorithm and TEB algorithm,it is found that the improved DWA algorithm has good dynamic obstacle avoidance effect compared with other algorithms,and the driving process is smoother,which has certain practical application value.