Considering Time-Varying Factor with Q Learning Strengthen Harmony Search Algorithm for Grid Task Scheduling
MA Shao-hu · Journal of Southwest China Normal University · 2015
In order to solve the shortcoming of the traditional factors of grid task scheduling algorithm,in which the time-varying price is not considered,the Q learning strengthen harmony search algorithm has beenpresented to do the grid task scheduling.Firstly,considering the time-varying price factors,the grid task scheduling model has beenimproved and a new scheduling model proposed;Secondly,the Q learning algorithm has beenused to improve the harmony search algorithm,whichhas been used for the scope search.And the Q learning algorithm has beenused for depth development,which balances wide-depth search ability of the algorithm;Finally,through the simulation results have beencompared with the similar algorithms show that,the algorithm and performance optimization model has better convergence speed,which has the optimal performance of a more comprehensive in two aspects of resource price satisfaction and task scheduling length.