A Chaos Harmony Search Algorithm for Solving Task Scheduling Problem

The paper studies a multi-objective task schedule problem since the scheduling problems involves multiple objectives in many practical cases. The cost profile and execution profile rather than the makespan are optimized altogether in distributed computing environment. The execution costs of the same task are heterogeneous and the cost of the task will decrease as its execution time is extended. A harmony search algorithm hybridized with chaos strategies is presented to solve the problem. Task model and the converting technique of the two-vector code make the harmony search (HS) algorithm adaptive to the task scheduling problem. Chaos strategies are designed to make sure the diversity of the solutions. The cost profile and execution profile of the task scheduling problem is evolved by harmony vector and search strategy. Self-adaptive cost scaling strategy is designed to effectively minimize the cost by extending the execution time according to obtained task slack and the overload of the processors. Empirical results on various benchmark instances validate the effectiveness and efficiency of our proposed algorithm.