In the context of industrial internet, industrial control systems are faced with the challenges of attack intrusion and vulnerability exposure. In order to ensure the safe and stable operation of industrial control systems, it is neccessary to explore the cause of system failures and clarify potential vulnerabilities. In this paper, based on the full-lifecycle security requirements and characteristics of industrial control systems, a multi-dimensional collaborative vulnerability analysis framework is proposed. With model-driven system static and dynamic vulnerability analysis, the system vulnerability knowledge is then generated. Meanwhile, the proposed framework firstly clarifies the meaning of vulnerability in industrial control systems, and the full-lifecycle coverage and integrated architectural features are benificial to realize the global vulnerability disclosure of industrial control systems.