Abstract: In order to solve the problem of not being able to accurately assess the transformer status during operation, this paper analyses the similarities, differences and changes in each stage of the entire transformer operation process based on big data processing and data mining technology through the massive data accumulated during the operation of power transformers, and derives the reliability factors and components that can reflect the operation of transformers.At the same time, by analyzing and selecting the characteristic gas content in the DGA data, combining with the service age of the transformer in the system as the key influencing factor for the evaluation of its operation status, the operation status characteristic analysis and risk analysis model of the transformer equipment are constructed, so as to achieve the prediction of transformer operating risk. The experimental results prove that the model proposed in this paper can accurately evaluate the operating status of transformers and predict the failure rate of transformers, and the results can provide analysis basis and decision-making guide for production personnel, operation and maintenance maintenance personnel and power grid dispatchers, which is of certain value to enhance the risk prevention capability of power grid and ensure the safe and stable operation of power system.