Research On Path Planning Algorithm Of Intelligent Search And Rescue Robot

In recent years,natural disasters and man-made accident have occurred frequently,such as earthquakes,tsunamis,large-scale fires and chemical gas leaks.When a disaster occurs,search and rescue personnel need to enter the disaster site urgently.At this time,the life and safety of search and rescue personnel are threatened,and the work efficiency is low.Therefore,intelligent search and rescue robot technology applied to post-disaster search and rescue work has become a current research hot spot of robotics.Path planning method of intelligent search and rescue robot is one of the key research contents in related fields.The key technologies of path planning mainly include two points: 1)Plan a path without collision,shortest distance,and high safety;2)How to obtain the environment information and build a model.For two aspects of path planning,the thesis first analyzes the intelligent path planning algorithm.The disadvantages of particle swarm algorithm in path planning are studied.It mainly includes that the particle swarm algorithm is prone to fall into a local optimum,fall into an endless loop and cause the path planning time to be too long.In order to address the above shortcomings,the thesis uses the following methods:1)the logarithmic function of adaptive adjustment decreases the inertia to change the weight;2)T Mutation factor in genetic algorithm to optimize individual speed values in particle swarm.By improving the particle swarm algorithm,the improved algorithm is compared with the basic particle swarm algorithm.Simulation experiments show that the rate of particle swarm optimization becomes faster,and it is closest to the optimal value.Secondly,the thesis analyzes and compares the current mature environmental models and their construction methods,and finally decides to apply the grid method as the construction method of the environmental model.As the map information in the search and rescue environment has been severely damaged,the thesis adds a rolling window method to local environment composition to perform environmental modeling.At this time,the grid size in the grid method is determined by comprehensive information such as the size of the obstacle,the diameter of the search and rescue robot,and the area of the obstacle in the total area.Based on this,the heuristic algorithm in the rolling window method is changed.In order to plan a collision-free,shortest path with the highest safety factor,a safety factor is added to the fitness function.Through simulation analysis,the proposed algorithm can more effectively plan the required path.In the end,the thesis introduces the improved algorithm into the search and rescue environment,and conducts test experiments.The algorithm is applied to a robot car,and the feasibility and effectiveness of the algorithm is verified by establishing a test scenario.By increasing the complexity of the search and rescue environment and testing the feasibility of the improved algorithm,the search time of the improved algorithm is reduced by an average of 45.9% compared to the comparison algorithm: the path length of the improved algorithm is reduced by an average of 10.9% compared to the comparison algorithm.It provides new ideas for path planning of search and rescue robots in a real-life disaster environment.