Inadequate network security defense measures threaten the information and property security of the state and the public, and how to safeguard network security is of vital practical significance. This paper proposes a dynamic security threat assessment model and a robust optimal control strategy to improve the efficiency of detecting network vulnerabilities and the accuracy of detecting network threats in a cloud computing environment. It also protects the privacy of the user's identity through a multi-factor continuous authentication method, encrypts and protects the user's data using a homomorphic encryption algorithm, and strengthens the ability of computers to resist intrusion. Three case studies are conducted to verify the effectiveness of the proposed technical approach: dynamic assessment of security threats, control policy, authentication, and network encryption. The results show that in the network vulnerability control policy and authentication case study, when adding the control policy Tk = 40, the percentage of network normal nodes, malicious nodes and restorer nodes are 0.98, 0.02, and 0.009, respectively, and the encrypted plaintext ASCII value is distributed in [0,60] in an unordered manner, and the network is in a very desirable security state. Real-time accurate assessments of network security state can be provided by cloud computing-based network security vulnerability and encryption technology.