The Research And Application Of The Parallel-Mixed Study On Frog Leaping Algorithm And Genetic Algorithm

As a heuristic intelligent optimization algorithm,the genetic algorithm has displayed its unique advantages in solving combinatorial optimization problems and complex discrete optimization problems.The genetic algorithm is also characterized by its high likelihood of fusing with other algorithms.Both numerical function optimization and combinatorial optimization are the classical research domains of the genetic algorithm,but this algorithm also has some shortages,especially when it comes to large-scale,high-dimension function problems and large city-scale traveling salesman problems.With high-dimension function optimization calculation and large traveling salesman problems as the starting points,this research conducts a mixed study on the genetic algorithm and frog leaping algorithm,aiming to solve massive data processing and calculation problems with mixed algorithm more efficiently.In order to achieve this goal,this research presents a number of constructive improvement proposals after making a series of multi-faceted research efforts.Firstly,relying on software technologies such as Tomcat,Redis and Spring Boot,etc.,this research proposes a simple and mature RESTful API parallel-distributed computing system scheme with RESTful API framework based on the complexity and problem characteristics of function optimization calculation and large traveling salesman problems.This solution is simple but mature,thus making the research focused more on mixed algorithm while serving the purpose.Secondly,in order to take full advantage of the local search capabilities of the frog leaping algorithm and to equip mixed algorithm with better global search capability,this research presents a discrete frog leaping search plan,which can realize self-adaptive frog leaping search based on the fitness difference of local optimal individuals and global optimal individuals and enable algorithm search to regulate the local and global search intensity automatically.Lastly,this research designs a frog leaping genetic mixed strategy.It analyzes the internal structure of mixed genetic algorithm and introduces improved discrete frog leaping algorithm on the basis of the genetic algorithm.Between the frog leaping search upgrading and genetic evolution operation,it realizes the effective combination of algorithms through the shared operation mode of local optimal individuals and global optimal individuals,which,combined with parallel-distributed computing schemes,will,in a way,further improve the overall global search capability and the local search capability of algorithm.According to theresults of the experiment and comparison of several universal calculating examples,the improved mixed genetic algorithm shows good search capability in solving traveling salesman problems.