| Smart power distribution and utilization is a crucial part of smart grid and is drawing more and more attention.With large-scale integration of distributed generations(DGs),controllable loads,electric vehicles(EVs),the smart distribution and utilization system(SDUS),which is comprised of heterogeneous software and hardware,is faced with the difficulties of heterogeneous big data storage and analysis.Therefore,the cloud computing technology,which has the capability of large-scale data processing,has become a key to fullfil the SDUS.This dissertation focuses on the research on the key cloud platform technologies for SDUS.The main contributions are as follows.1.A big data storage scheme is presented to meet the storage requirements of multi-source heterogeneous big data in SDUS.A detailed analysis of the composition of multi-source heterogeneous big data of the SDUS is introduced.Aiming at the storage realization of big data,a three-layer storage framework is then established,consisting of resource layer,metadata layer and actual data layer.Finally,using the techniques of Hadoop and MapReduce,a query scheme of big data is designed in SDUS platform considering three aspects of resources,storage and query,and provides a technological support for the storage and query of big data in SDUS.2.An optimization performance evaluation index system based on improved interval analytic hierarchy process(AHP)method is put forward to evaluate the performance of SDUS cloud platform.First of all,an optimization performance evaluation model of cloud platform is established from the five aspects of service providing ability,executive ability,network architecture capabilities,fast response storage capabilities and virtualization capabilities.Next,to deal with the shortage of analytic hierarchy process,the interval algorithm is introduced to express and cope with the uncertainty of expert scoring,and an improved interval analytic hierarchy process is proposed to solve the weight of cloud indice.Finally,the feasibility of the evaluation index and method of the SDUS cloud platform are verified using three cases to direct users to choose different SDUS cloud products for different purposes.3.The assessement methods for power users and household distributed photovoltaic projects based on SDUS cloud platform are proposed.In the evaluation of power users,a hybrid subjective and objective method for the index system is presented to consider the uncertainty of expert scoring and power users’ data at the same time.A hybrid power user evaluation method based on interval analytic hierarchy analysis and interval entropy method is proposed to achieve the objective evaluation of power customers and provide users the suitable services in smart grid.In the assessment of household distributed photovoltaic project,the uncertainty of expert scoring is considered.An evaluation method of households distributed photovoltaic projects based on interval balanced score card(BSC)is then put forward to provide a analystic tool for the development of household distributed photovoltaic in China.4.A cloud platform for SDUS is designed and verified.Firstly,a detailed SDUS cloud platform architecture is introduced by proposing three layers of system architecture,hardware architecture and application architecture,which achieves the top design of SDUS cloud platform.Then,from the three aspects of platform architecture,platform security,and policy adaptability,the cloud platform based user electricity utilization assessment and management system is realized,and provides a strong platform support for the following applications.Finally,a SDUS cloud platform is deployed in the laboratory using LuoyunCloud. |
PDF Full Text Request