数据呈爆炸性增长，导致存储设备的功耗和温度上升，硬盘的失效率急剧增加，系统的可靠性降低，如何在确保系统可靠性的同时，降低系统能耗，优化数据分布，成为混合存储系统急需解决的问题。从混合存储系统的层次结构上出发， 通过建立部件级的性能模型、功耗模型和温度模型，在满足工作温度门限的前提下， 根据不同应用特征和混合存储系统的异构组织方式， 研究基于温度受限的低能耗数据分布策略。 根据温度的变化特征、用户访问特征，保证在工作温度低于可控阈值的前提条件下，主动将数据聚集在较少工作集的存储节点上，形成高能效的自适应数据分布调度策略。通过结合现代的控制理论与混合存储系统的工作特点，研究混合存储系统中以模糊控制和模型预测为基础的温度控制理论，根据模糊逻辑动态的调整温度阀值和系统运行参数的控制方法，拟合模糊因子的方法， 建立精确混合存储系统的温度模型。该研究为提高混合存储系统可靠性、数据存储安全奠定理论基础。

With the explosive growth in data, power consumption and temperature of the storage devices are increased, which will raise hard disk's failure rate and lower the system's reliability. Therefore, it has become an urgent problem how to ensure system reliability, at the same time reduce system energy consumption and optimize data distribution for hybrid storage systems. Starting from the hierarchy of hybrid storage system, through the establishment of the performance, power consumption and temperature models at the component-level, under the premise that we can meet the working threshold of temperature, according to the different application characteristics and organization approaches of hybrid storage systems, we will conduct research on temperature-constrained power-aware data distribution strategy. According to the variation of the temperature and user access patterns, under the prerequisite that the working temperature is below controllable threshold, data access will be concentrated on a smaller working set of storage nodes, thereby forming the energy-efficient adaptive data distribution scheduling policy. Through the combination of modern control theory and hybrid storage system characteristics, we will study temperature control theory for hybrid storage system based on fuzzy control and model predictions. According to the control method that dynamically adjust temperature thresholds and control system's operating parameters through fuzzy logic, by fitting fuzzy factors, we will establish accurate temperature model for hybrid storage systems. This research will lay the theoretical foundation for improving the hybrid storage system's reliability and data security.