微型电脑应用

针对电力需求侧资源多元化、图片和关键词检索过程的召回率较低问题，提出融合聚类下的电力需求侧资源信息分布式多模块检索模型。采集包含互联网、相互关系、文件数据以及其他数据的需求侧资源信息，经结构化和非结构化处理后存储到本地。处理模块利用模糊c-均值融合聚类算法融合聚类采集模块获取的电力需求侧资源信息，经通信模块传输至检索模块后，采用查询扩展和分类相结合的信息检索算法，通过求解信息和查询式的相似度值获取检索结果，并通过显示模块呈现给用户。测试结果表明：所提模型的戴维森堡丁指数(DBI)高于1.5,融合聚类效果较好，可有效检索出所需电力需求测资源信息，且检索误差最大值为0.94%,检索误差较小，能够提升电力需求侧资源信息检索效果。

Aiming at the diversification of power demand side resources and the low recall rate of image and keyword retrieval, a distributed multi-module retrieval model of power demand side resource information under fusion clustering is proposed. The demand side resource information including Internet, interrelationship, file data and other data is collected, and stored locally after structured and unstructured processing. The processing module uses the fuzzy c-means fusion clustering algorithm to fuse the power demand side resource information obtained by the clustering acquisition module after being transmitted to the retrieval module through the communication module, it uses the information retrieval algorithm combining query expansion and classification to obtain the retrieval results by solving the similarity value of information and query, and presents them to users through the display module. The test results show that the Dawies Bouldin index(DBI) of the model is higher than 1.5, the fusion clustering effect is good, which can effectively retrieve the required power demand measurement resource information. The maximum retrieval error is 0.94%, and the retrieval error is small, which improves the retrieval effect of the power demand side resource information.