Aiming At Smart Grid Active Power Optimization Study Based On Cloud Computing

The State Grid Drawn up the world’s first programmatic smart grid standard which is used to guide how to develop smart grid----《strong smart grid technology standard system planning》, the standard put forward the technology standards roadmap for building a strong smart grid. Smart grid give overall consideration in power system on fields as power generation, transmission, distribution, substation, power and electric power dispatching and communication, developing the State Grid intelligently becomes an irresistible trend. In the aspect of power generation, so active optimization study is imperative to study further. meanwhile with the expansion of the power system scale, the requirement of computing and storage capacity increasing rapidly, it is necessary to introduce cloud computing to solve this problem,then the exist active optimization algorithm can’t adapt the requirement of parallel CAGA algorithm. To solve above difficulties, this paper mainly puts forward the following work:(1) In this paper, according to the IEC61970 standard, use the CIM model of Bus—Branch model express the connection relational data model for components which involved in active optimization, through the global topology analysis and local topology analysis to analyze the power grid. All above steps are laying the foundation to turn the original power resources data into a unified CIM/XML language.(2) We considering the existing active optimization algorithm may not be able to adapt to the characteristics of smart grid such as big data and real-time response in the future, this paper studies the cloud adaptive genetic algorithm, analyzes its advantages and disadvantages, finally according to its shortage as: slow convergence speed, easily plunged into local optimal solution and so on, designed an improved CAGA algorithm and applied it to the active optimization calculation.(3) We transplant the improved cloud adaptive genetic algorithm to the Hadoop, then based on cloud platform finish the improved CAGA algorithm parallel design, meanwhile design implementation steps of each functional part. Then, we build Hadoop cloud platform, through the JAVA programming algorithm, embedded the CAGA algorithm and improved CAGA algorithm in a Hadoop pseudo-distributed mode and verified algorithms by calculating power system examples, finally discusses the superiority of the improved CAGA calculate.