基于HACCP计划的水产品区块链溯源查询优化研究

Query Optimization Research for Blockchain-based Aquatic Product Traceability with HACCP

摘要:

结合危害分析与关键控制点（Hazard analysis and critical control points，HACCP）计划的溯源系统因使用区块链提升了溯源数据可信度，已成为水产品质量安全风险溯源中必要的技术架构。区块链在解决了溯源可信的同时也降低了查询效率，优化区块链系统提升查询效率是一个亟待解决的问题。本文以巴氏杀菌蟹肉为例，通过分析巴氏杀菌蟹肉HACCP计划质量安全风险的数据特点，结合实际业务场景，根据数据风险等级构建低风险数据链下存储，高风险数据根据最终溯源需求的不同分别构建映射数据上链和细节数据上链的"链下数据库+双链"数据存储模式，实现数据有序分流，提高查询效率;在智能合约中集成布隆过滤器，避免对误查询做全链遍历，提高区块链查询效率。最后基于Hyperledger Fabric环境进行测试，结果表明，"链下数据库+双链"数据存储模式优于传统单链存储模式，在总数量达10 000条时，查询效率为88%;集成布隆过滤器的查询合约相较传统智能合约效率提高15%。优化改进后区块链系统吞吐量仍能达305笔/s，满足区块链溯源系统要求。

Abstract:

The traceability system integrated with the hazard analysis and critical control point (HACCP) plan enhanced the credibility of traceability data by adopting blockchain technology, and became an essential technical architecture for quality and safety risk traceability of aquatic products. While blockchain ensured traceability data credibility, it also reduced query efficiency. Optimizing the query efficiency of blockchain systems was identified as an urgent problem to be solved. With pasteurized crab meat as the research object, the data characteristics of quality and safety risks in the HACCP plan for pasteurized crab meat were analyzed combined with actual business scenarios. According to data risk levels, low-risk data was stored off-chain;for high-risk data, mapped data and detailed data were respectively stored on separate chains based on differentiated end traceability requirements. Thus an off-chain database + dual-chain data storage model was constructed to realize orderly data shunting and improve query efficiency. Additionally, bloom filters were integrated into smart contracts to avoid full-chain traversal for invalid queries and further boost blockchain query efficiency. Experiments were finally conducted based on the hyperledger fabric environment. The results showed that the proposed off-chain database + dual-chain storage model outperformed the traditional single-chain storage model. When the total data volume reached 10,000 entries, the query efficiency was improved by approximately 88%. The query smart contract integrated with Bloom filters achieved a 15% efficiency improvement compared with traditional smart contracts. Meanwhile, the throughput of the optimized blockchain system remained at 305 transactions per second, meeting the operational requirements of blockchain traceability systems.