Research And Application Of Dual Population Ant Colony Algorithm Based On Heuristic Mechanism

Mobile robots are an important means to improve productivity,liberate labor,and improve people's living standards.Therefore,how to maximize the intelligence of mobile robots and realize autonomous navigation and control is deeply concerned by researchers at home and abroad.As one of the key technologies for mobile robot autonomous navigation,path planning has become a hot topic at home and abroad.Since its introduction,the ant colony algorithm has achieved a lot of research results in solving the path planning problem.This paper mainly studies the robot path planning problem based on ant colony algorithm.This paper analyzes the similarities and differences between the four classical improved ant colony algorithm and AS algorithm from the perspective of pheromone update method,and also analyzes the relationship between the parameters of ant colony algorithm,including basic parameters,heuristic information and pheromone update related parameters.From the analysis,it is concluded that the construction of heuristic functions and the update of pheromones play an important role in improving the performance of the algorithm.The specific tasks are as follows:Firstly,aiming at the problem that the traditional ant colony algorithm is easy to fall into local optimum and the convergence speed is slow,an improved ant colony algorithm with direction information is proposed.The specific improvement is based on ACS algorithm and A* algorithm.The direction information is not introduced in the early stage of the search to ensure the diversity of solutions.The target point information is introduced into the heuristic function of the ACS in the path back range,and the effect of the target point information on the ant action is adjusted by the dynamically changing weight coefficient.Simulation experiments show that the improved algorithm not only accelerates the convergence speed,but also improves the quality of the solution to some extent.Secondly,this paper presents a heuristic mechanism ant colony algorithm based on heuristic function construction and pheromone update.For the existence of concave obstacles at the starting point or the end point,the algorithm will fall into the local optimum problem.Based on the ACS algorithm,two aspects are improved.In the heuristic construction,the distances from the candidate node to the end point are sequentially sorted to assign weights.That is,the closer the next node is to the end point,the larger the distribution weight is,and conversely,the smaller the weight is assigned.In the aspect of pheromone update,the penalty function is introduced in the following two situations.One is that the ant cannot reach the end point,that is,the ant enters the death state,and the other is that the length of the current optimal path has not changed for 10 consecutive generations.The experimental results show that the improved algorithm based on heuristic mechanism has good performance in convergence speed and quality of solution.It can be used not only in general obstacle environment,but also in special obstacle environment,and its applicability is better.Finally,an improved dual-population ant colony algorithm and its application in robot path planning are proposed.In the construction of heuristic function,the distance between each node and the target point in each direction is calculated,and then the running direction is judged.In the aspect of pheromone update,the information interaction condition is set,and the pheromone of the two populations is exchanged when the setting condition is satisfied.The improved double ant colony algorithm is mainly to solve the path planning problem in large-scale environment.The large-scale map acquisition method is to convert the laser sensor pgm format map into a grid map.The scale of the converted grid map is 354×354.With the help of the Gazebo simulation platform,the robot path planning is simulated in the simulation environment.The keyboard control robot completes the scan work to save the map,calls the map to set the start point and the end point,and verifies the simulation effect of the robot path plan.Verify the actual path planning of the robot by operating the Turtle Bot2 robot.