Operators in the main control room of a nuclear power plant have a crucial role in supervising all operations, and any human error can be fatal. By providing operators with information regarding the future trends of plant safety-critical parameters based on their actions, human errors can be detected and prevented in a timely manner. This paper proposed a Sequence-to-Sequence (Seq2Seq)-based Long Short-Term Memory (LSTM) model to predict safety-critical parameters and their future trends. The PCTran was used to extract data for four typical faults and fault levels, and eighty-six parameters were selected as characteristic quantities. The training, validation, and testing sets were collected in a ratio of 13:3:1, and appropriate hyperparameters were used to construct the Seq2Seq neural network. Compared with conventional deep learning models, the results indicated that the proposed model could successfully solve the complex problem of the trend estimation of key system parameters under the influence of operator action factors in multiple abnormal operating conditions. It is believed that the proposed model can help operators reduce the risk of human-caused errors and diagnose potential accidents.

The introduction of digital technologies into the main control room of a nuclear power plant also introduces new human errors. The operator log records the control information of operators on systems and equipment, and provides an important data source for the retrospective investigation of operating events in a nuclear power plant. A traditional operator log review is conducted manually, which has some major problems, such as being time-consuming and inefficient. This paper proposes an automatic detection method for operator logs, which models an operating procedure at three levels, including procedure, step and action. Such a model clarifies the overall logic and basic attributes of the operating procedure, and can be used as a standardized template of a control action sequence to compare with the actual operation actions in the operator log, so as to identify possible human performance deviations. This paper explains the method, and discusses the advantages and limitations of the proposed method.