Path Plannng Algorithms Of Mobile Robot In Dynamic Environment

Nowadays,mobile robots have been widely used in various fields.As an important part of mobile robot technology,path planning has become a research hotspot.Robot path planning methods can be divided into global path planning and local path planning according to different environmental information.Global path planning algorithms always have poor real-time performance and local path planning algorithms are easy to fall into the local minimum due to the lack of global prior information.In view of the shortcomings of the above two kind of algorithms,a hybrid path planning algorithm based on the neural network and rolling windows is proposed to complete the path planning of mobile robots in dynamic uncertain environments.The principle of the global path planning algorithm based on neural network and the local path planning algorithm based on rolling windows is introduced and the simulation experiments are completed in Matlab.The core idea of the global path planning algorithm based on neural network is to measure the path proficiency by establishing the penalty function,and transform the path planning problem into a function extremum optimization problem.The reason why the original algorithm is easy to fall into the local minimum value is discussed and an improved algorithm based on simulated annealing is proposed.Based on the consideration of algorithm efficiency,the momentum gradient descent method and the path point detection mechanism are added to accelerate the efficiency of the improved algorithm.The local path planning algorithm based on the rolling windows draws on the predictive control theory,and selects the sub-target points in the current rolling window through heuristic search to complete the whole path planning process.The defect of the original algorithm that can't escape the concave obstacles is discussed and the pattern along the wall is introduced to escape the concave obstacles.Based on the above two algorithms,a hybrid path planning algorithm which uses the inflection point to split the global optimal path is proposed finally.The hybrid path planning algorithm uses the rolling windows to make the mobile robot reach each sub-target point separately so as to complete the whole path planning process.The relevant simulation experiments are carried out in Matlab,and the effectiveness of the algorithm is verified.Based on the three different points bewteen the Gazebo and the Matlab,the algorithm has been improved.With the Husky as the simulation object,the improved algorithm program releases the speed command to control the motion of Husky by subscribing to the lidar and odometer-related sensor messages released by the Husky.Husky can successfully avoid unknown obstacles and successfully complete path planning in a dynamic and uncertain environment at last.