关键词: 区块链, 源头数据, 云模型, 共识机制, 复杂动态赋权算法

Abstract: The blockchain with its features of data integrity and traceability ensures the authenticity and reliability of the data once it is recorded in the agricultural product traceability system.Currently，no mechanism is available to verify the reliability of data before it is uploaded to the blockchain，which may result in malicious modifications and system errors to the data uploaded to the blockchain.In response，this paper proposes a design of a blockchain-based origin data verification mechanism.Multidimensional data cross-validation algorithms and multisource data matching calculation algorithms are employed to verify the origin data of agricultural product production and logistics；a customer evaluation assessment algorithm based on cloud models and complex dynamic weighting is introduced to validate customer transaction evaluation data in the trading process and calculate genuine evaluation data；a credit rating is assigned based on the data validation mechanism’s results for node data and a reputation-based consensus mechanism CPOS（Credit Proof of Stake），is designed based on POS（Proof of Stake）consensus，making it more challenging for nodes with more malicious behaviors to obtain bookkeeping rights.This approach resolves the unfair competition for bookkeeping rights caused by adding coin age values in POS consensus and curbs cheating behaviors by nodes to some extent，ensuring the authenticity of origin data.Our experiments show this mechanism effectively filters out malicious and false origin data，maintaining the source data in agricultural product traceability systems within a relatively reasonable range.