[目的/意义] 针对传统数字资源保存方法存在的单点故障、隐私泄露及保存能力不足等问题，提出基于区块链与分布式存储的融合模型，以解决数据易篡改、权限管理低效和跨机构协作信任缺失等核心挑战。 [方法/过程] 设计分层架构模型，结合区块链存证元数据和IPFS分块存储文件，利用智能合约实现资源上传、权限控制与数据修复的自动化；通过AES加密保障隐私安全，并基于哈希冲突概率计算（4.26×10^-27）和容灾能力量化（文件年可恢复率99.99%）验证模型可靠性。 [结果/结论] 模型在数据不可篡改性、隐私保护（AES-256破解需3.68×10⁵¹年）和容灾能力上显著优于其他方法，适用于数字档案、医疗数据等高安全性场景。提出分层存储与数据去重压缩等策略，优化系统性能与成本。模块化设计降低开发复杂度，托管服务集成提升部署效率。研究表明，区块链与IPFS协同机制可以为数字资源保存提供可靠路径，未来需进一步平衡性能与成本以支持规模化应用。

[Purpose/Significance] Addressing issues such as single point of failure, privacy leakage, and insufficient ability of traditional digital resource preservation methods, an integrated model based on blockchain and distributed storage is proposed to solve the core challenges such as easy data tampering, inefficient permission management and a lack of trust in cross-institutional collaboration. [Method/Process] This study designed a hierarchical architecture model that integrates blockchain depositary metadata and IPFS chunked storage files. And smart contracts were utilized to automate processes including resource upload, permission control and data recovery. Data privacy was safeguarded through AES encryption. The model's reliability was rigorously verified based on a hash collision probability calculation (4.26×10^-27) and a quantified disaster tolerance capacity quantification (annual file recoverability rate of 99.99%). [Result/Conclusion] The model significantly outperforms other methods in data immutability, privacy protection (it takes 3.68×10⁵¹ years to crack AES-256) and disaster tolerance capacity. And it is suitable for high-security scenarios such as digital archives and medical data management. The study also proposes strategies such as hierarchical storage, data deduplication and compression to optimize system performance and reduce cost. A modular design reduces development complexity, while the integration of managed services enhances deployment efficiency. This research shows that the collaborative mechanism of blockchain and IPFS provides a reliable path for digital resources preservation. In the future, it is necessary to further balance performance and cost to support large-scale applications.