针对传统数据安全防护方法不足，提出基于国密算法的“云计算+大中台”防护方案。该方案融合智能体决策模型认证身份，确保合法访问。数据采用分组加密与复合变换强化加密效果。引入中央授权机构生成密钥与策略，结合密钥管理，实现双重防护。实验验证，该方法在抵御恶意攻击上表现卓越，远超常规方法。尤其在恶意节点密集环境下，通过多次迭代防护，数据传输的全局信任值显著提升，从0.58跃升至0.83，彰显了其高效、稳定的防护能力，为云计算及大中台数据安全提供了坚实保障。

Aiming at the shortcomings of traditional data security protection methods,a cloud computing+large and medium-sized platform protection scheme based on national security algorithms is proposed. This solution integrates intelligent decision-making models to authenticate identities and ensure legitimate access. The data is encrypted using group encryption and composite transformation to enhance the encryption effect. Introducing a central authorization agency to generate keys and policies, combined with key management, to achieve dual protection.Experimental verification shows that this method performs excellently in resisting malicious attacks,far exceeding conventional methods. Especially in environments with dense malicious nodes,through multiple iterations of protection,the global trust value of data transmission has significantly increased,jumping from 0. 58 to 0. 83,demonstrating its efficient and stable protection capabilities,providing a solid guarantee for cloud computing and data security in large and medium-sized platforms.