Abstract
Transient statistics is an important method for nuclear power plants to monitor the stress status and life of primary circuit equipment and pipelines, and to prevent the strength damage and fatigue damage of the primary circuit pressure boundary. The automatic identification technology of nuclear power plant transients developed in recent years helps to understand the state of nuclear power plants and take corresponding actions in time to ensure the safe operation of nuclear power plants. High-quality, accurate transient monitoring data is the basis for transient statistics. In order to improve the quality of transient data, the temperature and pressure data of domestic CPR1000 pressurized water reactor units over the years of transient monitoring were statistically and analyzed, and the common abnormal data forms were determined as “data missing” and “data jumping”. For abnormal transient data, the mean repair method, the interpolation repair method and the repair method based on object characteristics are proposed. The actual case analysis of the transient monitoring data of the domestic CPR1000 pressurized water reactor unit shows that the above method can effectively repair the abnormal points of the transient data, thereby significantly improving the quality of the data, and laying the foundation for the automation of the transient statistics of nuclear power plants.
