The Application Of Improved Ant Colony Algorithm In Path Planning

Some of the behaviors of living things in nature may be simply repetitive ways of living,but these tiny discoveries can often lead to great wisdom in the field o f intelligence.The ant colony Algorithms(Ant Colony Optimization,ACO)are one example of such wisdom.As a typical heuristic search algorithm,A CO may not have significant advantages compared with other algorithms in solving simple problems,however,it shows high efficiency in solving the NP-hard problems that the tradit ional optimization method is difficult to work.Therefore,the application of ACO not only reduces the cost of obtaining the solution of large-scale combinatorial optimizatio n problem,but also has the advantages of high speed and high precision.From the traveler problem to the robot path planning,from data mining to deep learning,there are figures of ant colony algorithm.Although ACO has many advantages,it also has its inherent defects as heuristic algorithm: slow convergence speed,premature stagnation,and so on.Aiming at resolving these problems,this paper designs a new type of ant colony algorithm based on search concentration degree,dynamic pheromone update and pheromone rollback mechanism.By introducing the "search concentration" factor into the selection strategy,the algorithm can adapt the range of the ant selectable city,that is,when the ant is too concentrated in the previous search iteration,it expands the search scope,and vice versa,thus speeding up the convergence speed.In addition,the dynamic update method of pheromone can make full use of the current obtained solution and adjust the pheromone increment on the path,so that the improved algorithm can better search the global optimal solution.Finally,the participation of pheromone rollback mechanism makes it easier for the algorithm to jump out of loca l extremum to avoid the occurrence of premature stagnation.The simulation results show that the improved algorithm has better performance in convergence speed,solution accuracy,and stabilit y,compared with the basic ant colony algorithm and several other ant colony algorithms that are involved dynamic update method.In addit ion,the improved ant colony algorithm is also applied to the tourism route planning in the scenic area,and a multi-objective route planning model of scenic spot is designed,which u ses different methods according to the different season of tourism,that is,the multi-objective route planning model is used in the busy season,and the improved ant colony algorithm is used to plan the path in off season.The advantage of this route planning model is that it does not use the path length alone as the target of the road selection,but takes into account the influence of populatio n density on walking time when the number of tourists is high,the effect of the bearing capacit y of scenic spots on the wait ing time,and the personal interests of tourists.Taking the scenic spot of "Yunnan National Village" as an example,using the method of combining real data with random data,the route planning model is compared wit h the improved ant colony algorithm under the same conditions,and the result show that the new model can effectively relieve the excessive accumulation of tourists,shorten the walking and waiting time of tourists,thus improving the satisfaction of tourists.