摘要
针对电力工程建设数据规模日益庞大,而现有数据处理方法无法满足需要的问题,文中基于B/S结构从客户端、服务器端和数据采集端3部分构建了基于深度学习与聚类算法的电力工程建设数据分析系统。其分析了用户管理、用户功能和信息管理3个模块的主要功能与作用,提出了基于深度学习与聚类算法的工程造价预测功能的实现方法。测试结果表明,所设计系统的各项功能模块均能够正常工作,可以满足用户功能需求。系统中基于深度学习与聚类算法的工程造价预测方法,通过相同特性数据的聚类分析及关系模式的自动提取学习,其预测误差均在10%以内。相比仅采用深度学习算法或BP神经网络的传统算法,具有较高的预测准确性。
Aiming at the problem that the data of electric power engineering construction is increasingly huge,but the existing data processing methods can not meet the needs.Based on the B/S structure,this paper constructs the data analysis system of electric power engineering construction based on deep learning and clustering algorithm from three parts:client,server and data acquisition.The main functions of user management,user function and information management are analyzed,and the realization method of engineering cost prediction function based on deep learning and clustering algorithm is proposed.The test results show that all functional modules of the designed system can work normally and meet the user’s functional requirements.In the system,the engineering cost prediction method based on the deep learning and clustering algorithm,through the clustering analysis of the same characteristic data and the automatic extraction and learning of the relationship mode,the prediction error is within 10%.Compared with the traditional algorithm only using the deep learning algorithm or BP neural network,it has higher prediction accuracy.
作者
刘沁
LIU Qin(State Grid Fujian Electric Power Co.,Ltd.,Fuzhou 350000,China)
出处
《电子设计工程》
2021年第3期27-30,35,共5页
Electronic Design Engineering
基金
国家电网科技项目(2018BR3677)。
关键词
深度学习
聚类算法
工程造价预测
系统设计
deep learning
clustering algorithm
engineering cost prediction
system design
