传统网格计算的信任管理模型中,网格服务提供者（GSP）难以有效评估和管理数量庞大的用户的信任度,且难以抵御Whitewash攻击并保障自身安全.为此,文中提出一种基于云模型的网格用户主观信任管理模型.该模型由自治域信任代理（UTA）管理域内用户的信任度,GSP则利用云模型评估UTA.实验结果表明,该模型符合网格架构的结构特点和信任的模糊、随机属性,克服了传统信任推荐方法的缺点,使GSP对网格用户的信任评估更加准确、合理.

In the traditional trust management model of grid computing,it is difficult for grid service providers（GSP） to effectively evaluate and manage the trust degree of a huge number of grid users,and to resist Whitewash attack and ensure the safety.In order to solve these problems,a cloud model-based subjective trust management model for grid users is proposed.In this model,the trust degree of users in the grid domain is managed by the user trust agent（UTA）,and GSP only needs to evaluate the trust of UTA with the help of cloud model.Experimental results show that the proposed management model,in accordance with grid architecture as well as with the fuzziness and uncertainty of trust,overcomes the disadvantages of the traditional trust recommendation method and makes the evaluation of trust more accurate and reasonable.