| Blockchain is a decentralized infrastructure and distributed computing paradigm extending from bitcoin to finance,the Internet of things,energy trading and other fields.Micro-grid is a distributed small-scale power generation and distribution system,which has realized the circular flow of electricity through adjacent distributed energy trading according to the different needs of prosumers.In micro-grid,blockchain can break the bottleneck of lack of trust between participants to reach a consensus and promote energy trading systems to realize decentralization and autonomy.However,although blockchain technology can effectively solve these problems including lack of trust,malicious tampering and false transactions,the open and transparent characteristics of the blockchain make the distributed energy trading model based on the blockchain extremely vulnerable and attacked,leading to the disclosure of user’s privacy.For this reason,this thesis combines differential privacy algorithms to design relevant privacy-preserving schemes for different stages of distributed energy trading.(1)At the trading matching stage,to develop optimal pricing and trading strategies to match the right parties for the transaction,this thesis proposes a double-sealed bid auction scheme.Except meets key economic properties,the scheme will determine the parties for the transaction based on the users’ quotations,bids,volumes,average price and other factors.In the meanwhile,to protect the privacy of users in the auction process,this thesis proposes the BDP(Blockchain-based Differential Privacy)algorithm based on differential privacy theory.Finally,this thesis combines the BDP algorithm with the double-sealed bid auction scheme to achieve a secure and efficient double energy auction privacy-preserving scheme,which protects the privacy of energy auction data while achieving optimal price and transaction strategy.Finally,the energy auction data privacy-preserving scheme is implemented on the Fabric consortium blockchain platform.this thesis analyzes the impact of the BDP algorithm on the auction data through experiments and proves that the BDP algorithm satisfies both differential privacy requirements and mean effectiveness through privacy analysis and data validity analysis.The effectiveness of the energy auction data privacy-preserving scheme in terms of average effectiveness,user satisfaction and social welfare is also demonstrated.(2)At the trading payment stage,to protect the privacy of trading data,this thesis proposes a privacy-preserving scheme based on account mapping technology and local differential privacy algorithm.The scheme redesigns the data perturbation mechanism of differential privacy based on Laplace distribution to make it suitable for blockchain technology and constructs the BLDP(Blockchain Local Differential Privacy)algorithm based on this perturbation mechanism to preserve the privacy of trading data on the blockchain.In the meanwhile,to ensure the correctness of transactions and to hide the characteristics of the transaction curve,the scheme uses account mapping technology to realize the association between users and multiple accounts,then designs an exponential smoothing prediction algorithm to calculate the transaction prediction value of each account,and finally uses BLDP algorithm to perturb the transaction prediction value to obtain the real transaction value and trade.Finally,the energy trading data privacy-preseving scheme is implemented on the Fabric consortium blockchain platform.this thesis analyzes the impact of the BLDP algorithm on trading data based on experiments and proves that the BLDP algorithm satisfies the differential privacy requirements through privacy analysis.The feasibility of energy trading data privacy-preserving scheme in preventing link attacks,differential attacks and data mining is also demonstrated. |
