Research Of Scheduling Algorithm Based On Hybrid Adaptive Genetic Algorithm In Computing Grid

With the development of the Internet, the technology of Grid Computing has come into being in recent years, which has the ability of changing the way the world works as well as internet. Grid Computing System and its key techniques include grid nodes, Wideband network system, resource management and task scheduling tools, monitoring tools and visualization tools of application layer etc. Among these technologies, the performance of task scheduling technology directly influences the efficiency and quality in task scheduling. So the task scheduling problem has become a problem to be faced and solved during the research and application of the grid.In this paper, based on the analysis of the challenges that traditional task scheduling algorithm , we propose a dynamic scheduling algorithms under the pattern of Adaptive Hybrid Genetic Algorithm, and carry on the simulation and confirmation through MATLAB. The concrete contents include:(1) Understand the present status and future developments in some areas of grid technology at home and abroad, analyze and compare the implementation and existing issues in grid task scheduling algorithms.(2) Analyse the existing problems in genetic algorithm. Adjust fitness function by the individual evolutionary, which prevents the"beceive"question in evolution process; utilizing clustering algorithm to implement population division, making each sub-population evolve alone, which can improve the algorithm convergence and the rate of convergence; In addition, adjust crossover and mutation probability, which making the genetic operation more approach to real environment, avoiding premature convergence .(3) In view of the existing problem in task scheduling, trying to introduce the genetic algorithm into scheduling strategy, which can make resources processor take the shortest time to complete task and solve the allotment problem between the resource in dispatcher process .(4) In order to test and verify the optimization performance of the presented algorithm, carried on the simulation and confirmation through MATLAB, and then compared with the result of the other genetic algorithms.