Abstract: Human-caused failure is the main factor causing nuclear power accidents. In construction, subjective human-caused failure can occur when the subjective behavioral will of the participants conflicts with the quality control regulations, so an intelligent quality witnessing method is proposed to prevent failure based on the subjective human-caused failure control structure. On the basis of analyzing the collaborative quality witnessing model and subjective human factors failure in nuclear power equipment, it construct a subjective human factors failure prevention safety control structure for collaborative quality witnessing based on the systems-theoretic accident model and processes method, identify the risk factors of subjective human-caused failure caused by the lack of safety constraints, and utilize the decision making trial and evaluation laboratory method to calculate the effect of key risk factors to reveal the technical and management barriers of subjective human-caused failure event prevention. Through the deep integration of IOT information sensing and big data technology, the collaborative quality intelligent witnessing method for subjective human-caused failure prevention is designed and its architecture is given. The developed intelligent witnessing platform is demonstrated to be well applied in the civil engineering of the CGNPC San'ao nuclear power project. The proposed collaborative quality intelligent witnessing method can provide reference and reference for the subjective human-caused failure prevention of complex products represented by nuclear power equipment.

Key words: nuclear power construction, collaborative quality witnessing, subjective human factors failure, intelligent manufacturing