To address the challenges of existing transmission network fault diagnosis methods in accurately characterizing fault evolution processes under complex operating conditions and often lacking sufficient interpretability of diagnostic results, this paper proposes a transmission network fault diagnosis and confidence assessment method based on a protection-component interdependent causal network. First, the physical topology of primary components and the causal relationships of secondary protection are analyzed, and a hierarchical heterogeneous causal network model is established. Second, to improve computational efficiency, a local network segmentation method is proposed based on the locality of protection actions and fault regions, and local causal subnetworks of candidate faulty components are extracted. Then, a confidence scoring model integrating protection action evidence, circuit breaker execution feedback, and key causal pattern matching is constructed to locate faulty components and infer complex conditions involving protection or circuit breaker failures, such as failure to operate and maloperation. Finally, multiple typical fault scenarios are constructed based on a five-bus test system, and comparative analyses are conducted with existing diagnosis methods. The results show that the proposed method can effectively identify primary component faults and abnormal protection actions in transmission networks, providing accurate and highly interpretable diagnostic results.