针对移动对象的频繁更新导致的I/O开销较大的问题,本文在密度格网索引中引入统一的磁盘管理机制,提出两个重要的批量优化方法,即批量写盘策略和预取.这些方法解决了移动对象更新过程中的I/O操作的无序性和离散性等问题,有效地提高了索引的更新性能.文中给出了这些优化方法的设计思想和算法,并分析了优化前后更新操作的时间成本变化.实验结果表明,我们的方法达到较高的更新效率且在某些方面显著优于现有的索引. In view of the large I / O cost caused by the frequent updating of mobile objects, this paper introduces a unified disk management mechanism in density grid index, and proposes two important batch optimization methods, batch write strategy and prefetch. The method solves the problems of disorder and discretization of I / O operations in the process of moving objects and effectively improves the index updating performance.The design ideas and algorithms of these optimization methods are given in this paper, The time cost of the update operation changes.The experimental results show that our method achieves a high update efficiency and in some respects is significantly better than the existing index.