托管型数据中心激励机制优化算法分析与研究

摘要

Due to the low efficiency of incentive mechanism in colocation data center while facing emergency demand response,the dynamic programming,genetic algorithm and particle swarm optimization and hybrid algorithm were exploited to address the optimization problem in incentive mechanism.With the features of emergency demand response and the demand of four algorithms,the optimization problem was elaborately formulated and it was proved that the problem belongs to an NP-complete problem.The simulation was conducted.The results demonstrate that four algorithms not only maximize the supplied electric power for tenants,but also show the effectiveness and efficiency in addressing the problem.%针对紧急需求响应下托管型数据中心激励机制中胜标方决策效率低的问题,分别采用动态规划、遗传算法、粒子群优化及混合算法等方法对其激励机制的优化展开分析与研究.根据紧急需求响应的特点和算法需求建立托管型数据中心的激励机制优化模型,证明激励机制中的胜标方决策问题为一个NPC问题.从理论上分析优化算法使用的可行性及复杂度性,为算法在激励机制中的应用提供了理论基础.通过实验仿真,分别从4种算法的处理规模、性能以及时间复杂度的角度进行阐述和对比,实验结果表明,4种算法优化了胜标方选择最大化可供电力并满足管理员的最大支付,体现了优化算法解决问题的有效性及高效率性,提高了决胜标方决策的效率.