基于B+树的电力大数据分布式索引

摘要

随着电力信息化的发展，电力数据来源广泛，具备体量大、类型多的特点，其中设备监测数据以及业务数据大多是浮点型、字符型数据，具有一定的时序性和结构化的特点。在数据检索时可能是对不同类型数据的联合查询，同时在大规模数据检索时存在查询效率不高，检索结果无法满足跨范围匹配的问题，对此本文设计并实现了一种基于B+树和倒排索引的分布式混合索引结构，引入层次化很合索引的思想，将数据集中的数据属性和属性值划分开来，并实现索引的并行化，提高了数据的索引构建时间和检索速度。%With the development of electric power information technology,electric power data comes from a va-riety of sources,and have the characteristics of large volume and a lot of types,equipment monitoring data and business data mostly floating-point type,character type,has the characteristics of sequential and structured. In the data retrieval,it is possible to combine the different types of data,the query efficiency is not high According to the characteristics of electric power data,this paper proposes a distributed hybrid index structure based on inverted index and B+-Tree,introducing the idea of hierarchical indexing,the data stream contains attributes and attribute values separated to improve the retrieval speed.