Research On Task Scheduling Technology Of An Improved Firefly Algorithm

The current society is in the era of highly integrated information.In order to process the large-scale and incremental user data,multi-core processors have become the mainstream of research and application in the field of computer hardware.The performance of the task scheduling algorithm is crucial to improve the computing resource utilization of multi-core processors.In order to maximize the advantages of the parallel structure of multi-core processors,it is necessary to develop a task scheduling algorithm with both high efficiency and stability,so as to achieve the purpose of shortening the task scheduling length while improving the resource utilization of multi-core processors.This paper deeply explores the characteristics of multi-core processor task scheduling problem and the optimization characteristics of standard firefly algorithm,and proposes an improved firefly algorithm IPFA and applies it to the field of multi-core processor task scheduling.The IPFA algorithm analyzes the shortcomings of the standard firefly algorithm through research and gives corresponding improvement strategies: aiming at the problems of uneven quality of fireflies in the population,uneven distribution of fireflies in the solution space and poor global coverage caused by the excessive randomness of the initialization method of the standard firefly algorithm,an initialization method based on the uniform division of the solution space and the relative attractiveness screening strategy is proposed,which not only ensures the diversity of individuals in the population,but also improves the quality of the initial population as a whole;in the position update formula of the algorithm,the speed term of the particle swarm algorithm is introduced to replace the original random interference term,and the experience of the speed term on the historical optimal position of fireflies is used to stably improve the algorithm’s ability to explore the solution space,thereby shortening the solution time of the algorithm;after the firefly population converges for the first time,the explosion operator of the fireworks algorithm is used to perform the explosion operation on the current optimal firefly individual,and the spark generated after the explosion is calculated and compared with the fitness value of the optimal firefly,so as to identify and jump out of the local optimal solution.Finally,combined with the improved strategy of the scheduling algorithm,a task scheduling model is constructed,the continuous solution space of the IPFA algorithm is mapped to the discrete solution set of task scheduling,and the corresponding relationship between the task scheduling length and the fitness function is established,and then get the IPFA multi-core processor task scheduling strategy.In order to verify the optimal performance of the IPFA algorithm in solving the task scheduling problem of multi-core processors,the simulation platform is used to compare the IPFA algorithm with the standard firefly algorithm,particle swarm algorithm,fireworks algorithm and their improved algorithms.The number of iterations and task execution time to compare the algorithm’s convergence speed and task scheduling length.Analysis of the experimental results shows that the IPFA algorithm has faster convergence speed and shorter task scheduling length,and the algorithm has both feasibility,efficiency and stability.