Research On Key Issues Of Power Transaction Data Governance Based On Blockchain And Deep Learning

With the rapid development of economy,the problem of global energy shortage and resource limitation gradually appears,and the traditional energy system has been severely challenged.The development of distributed new energy has become the trend of energy system research.Taking advantage of the value of distributed renewable energy can improve the reliability of electricity market.Combining conventional and renewable energy sources,the smart grid establishes a flexible but highly reliable,resilient,cost-effective and efficient electricity market network.Smart grids provide two-way energy flows and communications flows,integrate consumer decisions and provide a platform to manage energy flows and data flows.All kinds of power generation,electricity price and other data privacy content collected by smart grid are facing the threat of disclosure.Blockchain technology is a decentralized and transparent distributed data storage technology that can achieve secure and reliable data interaction and protect trading data.Blockchain technology is a decentralized,open and transparent distributed data storage technology that enables secure and reliable data interaction.It can protect transaction data.With the explosive growth of big data,the study of data quality management has become a new trend.In the electricity market,large amounts of heterogeneous multi-source data are generated by various edge devices,these data describe energy trading in the electricity spot market.Such trading data can be lost in transmission,which can result in wrong decisions.However,there is a lack of methods to describe and evaluate completeness of such trading data according to the underlying structure.Besides,there is also a lack of methods to repair such missing data according to the structure.In this paper,an architecture of blockchain and edge intelligence for trading data quality management is proposed,which can offload computing tasks from the edge devices to the edge servers to reduce the energy consumption and latency for edge computing.We consider the dependencies of the trading data,and propose an evaluation and repairing model for data completeness.To repair the missing data with security protected,a deep learning algorithm based on attention is designed,where blockchain is employed to trace low-quality data and repair missing values by a novel data quality based consensus mechanism.The numerical results show that the data completeness evaluation fully reflects the completeness of data in the electricity spot markets,and the repairing model improves the data quality.