关键词: 遗传算法, 并行任务调度, 任务向量矩阵, 加速进化策略

Abstract: Parallel task scheduling is NP-complete problem，which focuses on resource allocation and parallel task schedule，requiring high-performance scheduling algorithm and high-quality solutions.The paper presents a parallel task scheduling algorithm based on improved genetic algorithm，which introduces vector matrix to represent task，resource and scheduling relationship，and use heuristics when original colony is initialized，improving the quality of initial colony.And it adopts rule-bound crossover and mutation operation to improve individual quality.Besides，it proposes an evolution acceleration strategy to?avoid the premature effectively.Simulation result suggests that the algorithm can solve the parallel task scheduling problems effectively.

Key words: Genetic Algorithm（GA）, parallel task scheduling, task vector matrix, evolution acceleration strategy