Research And Implementation Of Data Analysis Based On SAP BI In Power Enterprise

As the market competition becomes increasingly intensified, information has beenbecoming the key factor for the enterprise to survive and develop. Nowadays, informationgrows explosively with business expansion and requires enterprises to improve theircapability of information processing. How to make full use of vast amounts of data, makinginformation truly playing a decision-making role and improving enterprises'overall andcore competitiveness, has become a critical issue for enterprises to face when they developto a certain stage. Business intelligence is the inevitable choice to solve these problems.SAP Business Intelligence is SAP's enterprise data warehouse and businessintelligence solutions, and also the core component of SAP next-generation technologyplatform SAP Netweaver. The advanced data warehousing functionality and businessintelligence platform technology of SAP Business Intelligence solutions can helpcompanies to complete business goals excellently. SAP BI can inherit, merge and convertthe related datasource of SAP application systems and all other external datasources, easilyintegrate with other business systems, change different data of heterogeneous systems intohomogeneous data in a unified data platform. Using online analytical processing (OLAP)technology as a structured data analysis methods, SAP BI can complete the group's allbusiness data processing and analysis.In order to enhance the company's business analysis capabilities, to create a unifiedbusiness analysis platform and to enhance the company's management level, large powerenterprise launched BI deepening project. Combined with the operational features of largepower enterprise, this paper analyzes the main difficulties that the large power enterprise isfacing:Large power enterprise has subsidiaries throughout the country, each of which hasvarious business systems but has no uniform data interfaces, so how to flexibly and fastlyextract data to a unified data platform has become a top priority .Large power enterprise has complex business structures, staff of different levels havedifferent business demands. How to implement structured data modeling that not onlysatisfies the demand of data analysis, but also take into account of the system performanceand storage requirements is a key factor.Different users tend to different display ways. How to flexibly display data in differentways to different end-users has become a prominent problem.This paper's focus and innovation are summarized as follows:Combined with the specific features of large power enterprise and the architecture of SAP BI, design the overall technical architecture, the overall data architecture and theoverall hardware deployment scheme of the large power enterprise, and also beginrequirements analysis, creating comceptual model and logical model. The overall technicalarchitecture enterprise data warehouse can be divided into three parts: the source systemlayer, data warehouse layer, presentation layer; the overall data architecture can be dividedinto four layers: data extraction layer, data processing layer, reporting data layer andpresentation laye; the overall hardware deploy scheme of large power enterprise requiresthat all the development work must be done on the development server, through thetransmission connection to transmit to the test server. If tests succeed, transmit the requestto the production server, otherwise modify in the development server.To collect data from a variety of heterogeneous systems into a unified data platform,and convert into the homogeneous data, and complete company-wide data analysis is thispaper's research priority. Large power enterprises'complex business scenarios determinethe diversity of source systems, mainly SAP source systems, external databases, text files,the corresponding interfaces are BI Service API, file interface, DB Connect interface. Allthe source systems must meet the BI's technical requirements.This paper combines with ETL(Extraction, Transformation,Load) data flow todesign and implement large power enterprise's data extraction. The source of ETL dataflow is the source system. Through BI Service API transport mechanisms, SAP sourcesystems can seamlessly integrate with SAP BI. SAP datasources can be defined in thesource system, BI Content (Business Content) predefined a number of standard datasources,and users can enhance them to meet their own needs. Additionally, users can also customizeSAP datasource using the table / view, the information set, SAP function module means tomeet their needs. Text source system is relatively simple, you can specify text file'sphysical path in the datasource to extract the text file on a file server. External database canintegrate with SAP BI using DB connect interface. In the external database you needdistribute access users for SAP BI and define views for all the data required; in BI, accessuser and view must be distributed to datasource. This approach ensures that BI systems canonly access the data in views that external database has defined for BI.After extracting dato into the unified data platform, the next is to consider how toimplement data modeling in SAP BI to meet the requirements for efficiency and ability touse existing models dealing with new demands. Master data is extracted from datasourcesand loaded directly to InfoObject after using transformation and data transmission process,serving for business data. Business data is extracted from operational datasource and loadedto write optimized ODS in data extraction layer without any change, achieving the realstorage in BI system. Data processing layer use standard ODS as main object to cleanse,integrate the data extracted from data extraction layer. Report data layer have unified dataanalysis models. InfoCube extracts data from standard ODS and forms aggregated data, inwhich users can analyze data through multi-angle. Multi-InfoProvider can contain standard ODS, InfoCube and InfoObject, achieving across data model analysis.BEx tools customize the display form of SAP BI data. They define fixed report anddevelop various applications and templates based on flexible queries. Data can be displayedin Microsoft Excel, also be issued with a JSP script WAD in the enterprise portal using avariety of flexible charts.In summary, combined with the operational features of large power enterprise, thispaper analyzes the main difficulties that the large power enterprise is facing. Design theoverall technical architecture, the overall data architecture and the overall hardwaredeployment scheme of the large power enterprise, and also begin requirements analysis,creating comceptual model and logical model. Extract data from a variety of heterogeneoussystems into SAP BI, and convert into the homogeneous data. Combined with ETL and theoverall data architecture, achieve data warehouse modeling hierarchically and display datain enterprise portal using BEx tools to customize display form to satisfy users'requirement.