Research On The Application Of Shortest Path Planning Algorithm Based On Multi-Factorial Optimization

With the rapid development of the national economy,the application scope of path planning continues to expand,and its main application fields include:logistics transportation path planning,robot obstacle avoidance path planning,drone path planning,etc.Therefore,algorithms for solving path planning problems have attracted the attention of many scholars.Traditional path planning algorithms are greatly limited by the large number of nodes and constraints,and the use of evolutionary algorithms to solve path planning problems has become a research hotspot in the field of artificial intelligence.However,in the case of limited computing resources,optimizing multiple path planning tasks at the same time can result in slower computational speed and reduced efficiency.The multi-factorial evolutionary framework proposed in recent years has shown good performance and potential in optimizing multiple tasks simultaneously.This article combines the multi-factorial evolutionary framework with the genetic algorithm in evolutionary algorithms to handle discrete multi task path planning problems.A topology dataset consisting of 300 cities was constructed,and a multi-factorial genetic algorithm framework for discrete path planning problems was designed.It can accelerate the optimization speed and improve iteration efficiency while performing multiple path optimization tasks simultaneously.A measurement method for the similarity between discrete tasks in the multifactorial genetic algorithm for discrete path planning problems is designed to address the characteristics of the similarity between tasks affecting the optimization results.A multi-factorial genetic algorithm for adaptive random mating probability in discrete path planning is proposed.The main research work of this article includes the following three aspects:1.A multi-factorial genetic algorithm for discrete path planning problems was proposed,achieving the goal of efficiently optimizing multiple discrete path planning tasks under the same computing resources.Firstly,design an algorithm running framework that combines multifactorial methods with genetic algorithms;Secondly,the elite genetic assortative mating method is designed to generate offspring population;Finally,a gaussian perturbation strategy was added to the endpoint of the path to avoid premature convergence and local optima,increasing diversity.In the experiment,numerical experiments and analysis are carried out on the shortest path problem model and the traveling salesman problem model respectively.Compared with other algorithms,the proposed algorithm is competitive and can speed up rate of convergence and improve the quality of the solution.2.A multi-factorial genetic algorithm for adaptive random mating probability in discrete path planning has been proposed,which has strong robustness to the evolution of tasks with different similarities.Firstly,based on the characteristics of discrete problems,calculate the distance and angle between task vectors;Secondly,a similarity measurement method for path planning tasks was designed using the calculated distance and angle;Finally,adaptively adjust the probability of random mating using similarity measures.In the experiment,numerical experiments and analysis are carried out on the shortest path problem model to verify the effectiveness of the proposed algorithm.3.Designed and implemented a path planning system for logistics companies,which can pre plan routes in actual logistics,analyze logistics costs,and manage personnel and logistics vehicles for logistics companies.