Path planning research is an important part of robotics technology and it is the key to mobile robots achieving autonomous mobility.Ant colony algorithm,as its name implies,is an algorithm that simulates the behavior of ant populations.The birth of the ant colony algorithm was firstly discovered by the Italian scholar M.Dorigo through the observation and analysis of the foraging behavior of the ant in nature,and was designed as a new type of bionic evolution algorithm.The ant colony algorithm is very good at solving discrete optimization problems with high time complexity.Based on its strong anti-jamming capability and easy parallel computing capability,the ant colony algorithm has been widely used in robot path planning,traveling salesman problems,task allocation,and traffic.Scheduling and other areas,and achieved good results.This paper is based on the static environment,using ant colony algorithm for path planning research.The basic principle of ant colony algorithm is introduced in detail.The mathematical model is established.The related concepts such as node transition probability,heuristic function and pheromone residual are defined.The improved version of basic ant colony algorithm is studied in depth.The grid method is used to model the environment of the robot's working space,and the ant colony algorithm model is partially improved to adapt to the path planning problem.The study found that the traditional ant colony algorithm has the disadvantages of slow convergence speed and easy to fall into local optimum when solving path planning problems,and the planned path has poor smoothness and low security,which is not conducive to the precise tracking control of the robot.This paper proposes an improved ant colony algorithm based on ant colony classification search strategy,designs a heuristic function that combines the neighbor heuristic factor and the goal heuristic factor,and classifies ant populations,and adopts different heuristic search strategies for each type of ant.A new path evaluation function is constructed,and the path turning angle parameter and the inflection point safety distance parameter are introduced to improve the smoothness and safety of the path.In order to further improve the performance of classified search ant colony algorithm,a multi-strategy-based classified search ant colony algorithm is designed,which includes a pheromone update rule based on dual feedback mechanism to improve the convergence speed of the algorithm,and a trap processing method based on the engineer ant strategy.In order to avoid the ant colony being trapped in the U-trap during the search process,it can slow down the operation efficiency of the algorithm,and based on the path backoff optimization strategy,the path secondary optimization method is adopted to avoid the local path oscillation problem of the planned path.The simulation results show that the classification search ant colony algorithm can plan a smoother and safer path,and the convergence speed of the algorithm is faster,which proves the effectiveness and feasibility of the improved algorithm. |