Research on Risk Calculation in Chemical Production Processes Based on Dynamic Bayesian Networks

In the chemical industry, the continuous innovation and development of cutting-edge technologies are driving the evolution of key industrial characteristics towards integrated production facilities, automated control systems, and digital information networking. However, the production processes in the chemical sector exhibit traits such as prolonged operation times, intricate process manipulations, complex and variable chemical reactions, and high-temperature and high-pressure production environments. When unfavorable factors such as internal disturbances or external abrupt changes affect the process system, the control over on-site production may be lost, potentially leading to accidents. Safety risk management in chemical production processes remains a significant challenge in the industrial domain. This paper proposes a methodology based on Dynamic Bayesian Networks (DBNs) for real-time calculation and assessment of residual risks in chemical production processes. This approach offers a robust scientific basis for making risk management decisions pertaining to typical chemical production processes in chemical enterprises.