Risk assessment of human-machine systems based on Hybrid Causal Logic model and Cognitive Reliability and Error Analysis Method

Accident evolution in human-machine systems is a complex process, often influenced by multiple factors and characterized by uncertainty. This paper proposes a risk assessment method for human-machine systems based on the Hybrid Causal Logic (HCL) model and Cognitive Reliability and Error Analysis Method (CREAM). The HCL model is a three-layer structure which incorporates Event Tree (ET), Fault Tree (FT), and Bayesian Network (BN). The ET-FT considers the impact of machine failures and human errors. In addition, this study also uses the BN to describe the common cause effect of Common Performance Conditions (CPCs) on multiple human error modes. Therefore, we employ CREAM to quantify human errors, and utilize a Binary Decision Diagram (BDD) algorithm to rank the importance measures of each event and calculate the accident probability of the risk scenario. We present a case study of a twin-engine aircraft experiencing a one-sided engine fire and apply the ET-FT-BN framework to model the risk scenario. We calculate the important measures and accident probability, thereby demonstrating the effectiveness of the proposed method. This paper provides a scientific and effective analysis tool for risk assessment of human-machine system.