Safety Evaluation Method of Guyed Tower based on Bayesian Network

This paper presents a data-driven safety assessment framework for transmission guyed towers by integrating Bayesian network (BN) with fault tree analysis (FTA) and field inspection data. The proposed model quantifies component failure probabilities through empirical data, eliminating subjective expert judgments in conditional probability table (CPT) construction. By establishing a bidirectional reasoning mechanism—forward propagation for risk prediction and backward inference for rootcause analysis—the framework systematically identifies critical failure factors under complex environmental and operational conditions. Case studies involving 29 high-voltage guyed towers in Hubei Province demonstrate that the method achieves a 92% consistency rate with actual failure records, providing a reliable basis for proactive maintenance strategies.