中文摘要：

云计算作为一种新兴的计算范式，越来越受到工业界和学术界的关注．云计算是一种商业性服务，当用户提交的实时应用因机器故障而没办法得到成功执行的时候，云提供商必须提供经济赔偿．提出一种新颖的云计算容错任务调度算法，算法通过采用主／副版本技术提供容错机制．同时，通过离散粒子群优化算法对截止期错失率、执行时间、执行成本、负载均衡度进行多目标优化．此外，为了提高系统资源的利用率，采用被动副版本重叠技术来减少冗余．实验结果表明该算法可以有效地减少任务失效个数，缩短任务总执行时间，减少任务总执行成本．

英文摘要：

As a new computing paradigm, cloud computing is receiving considerable attention in both industry and academia. Cloud computing is a commercial services. Cloud service provider must provide economic compensation when the real-time applications sub- mitted by end-users fail to execute for machine errors. In this paper, a novel fault-tolerant task scheduling algorithm for cloud compu- ting is proposed to solve the problem. The fault-tolerant mechanism implemented in the algorithm is supported by employing primary/ backup copy technique. Meanwhile, the discrete particle swarm optimization algorithm is adopted to optimize deadline missing ratio, execution time, execution cost and load balance. In addition ,to improve system resources utilization ,passive backup copy overlapping technique is employed to reduce redundancy. Experiment results indicate that the proposed algorithm can effectively reduce the number of failure tasks, shorten task execution time and cut down task execution cost.