Abstract: In order to improve the energy-saving level of task scheduling in heterogeneous computer systems, a multi-variable neighborhood energy-saving scheduling algorithm fused with pruning optimization is proposed. Processor constraint and time constraint neighborhood structures are constructed in the algorithm. The number of redundant processors is reduced by constraining the neighborhood of processors, thus lowering the overall energy consumption. Pruning optimization is introduced into the time and energy consumption neighborhood to improve the efficiency of local optimization. Simulation results show that the proposed algorithm achieves good energy-saving effect under different problem scales, processor capacity and communication ratio.

Key words: energy saving task scheduling, variable neighborhood structure, processor constraints, time constraint, pruning optimization