| AGV path planning plays a key role in the running process of AGV,although the research on AGV path planning has been around for decades,but due to AGV’s working environment is very complicated and obstacle avoidance path optimization is difficult,so there are still many places worth studying.The realization method is the key to the path planning in the process of AGV running,it plays an important role in the operation process of AGV path planning.in the process of path planning study for a long time,There are many path planning algorithm,it mainly divided into two categories,one is the traditional algorithm,the third is the intelligent algorithm that gradual rise of the recent years,But whether the traditional algorithm and intelligent algorithm,they have theirs advantages and disadvantages respectively,and the obtained results that using these methods is the most ideal not always,so it requires further research to do research.Based on the understanding of AGV related issues of the current’s questions,methods and existing problems that related to AGV path planning and taking into account the AGV operation in the real working environment,In this paper,the AGV path planning problem of single laser navigation is studied.The AGV path planning method that using the path planning of the common exists many disadvantages,it includes that length is not ideal and the smooth degree of path planning is not enough,so it is necessary to improve the existing path planning method,this paper mainly adopts two methods of path planning,one is the improved genetic algorithm,There are three main improvements in this algorithm,the first is to improve the initial population generation method,the second is the improvement of mutation operation,the third is the introduction of smooth operator,using the improved genetic algorithm is mainly because it has the characteristics of good parallelism and searching the multiple solution space.In this paper,We use improved genetic algorithm for path planning of single laser navigation AGV in the environment that exist number of different obstacles and different obstacle locations,by comparing with the basic algorithm to analyze the performance of improved algorithm,so as to explore the advantages and the disadvantages of single AGV navigation path planning methods that improvement of genetic algorithm and the basic genetic algorithm.Aiming at the disadvantage of the initial population fitness value is not high in the genetic algorithm,the paper adopts another AGV path planning method,which is called genetic ant colony algorithm.In this method,The initial population is not randomly generated but out of the ant colony algorithm iteration’s results in every generation,and then through the selection,crossover and mutation,smooth to get the final result,and after comparing with other algorithms shows that this algorithm has some improvement.Finally,according to the above two kinds of path planning algorithm,providing the corresponding MATLAB code for simulation and building a path planning simulation prototype system based on the C# and MATLAB software,Let AGV run in two often working environments using the prototype system planning,The results shows that the prototype system has a certain practical value. |
