An attempt to utilize functional modeling for enhanced application code analysis

Due to the complex and highly interdependent relationships among parameters in system programs, anomalies can lead to abnormal propagation, thereby degrading system performance and potentially inducing measurement or execution errors within the control system. Such errors can adversely affect the stability and safety of a nuclear reactor. To address this issue, this study introduces a function modeling-based analysis method for data anomaly traceback and parameter tuning assistance. This method enables root cause inference of anomalies when program outputs significantly deviate from expected values, facilitating further investigation. For parameter tuning, the method simplifies the direct model of the program through functional combinations, elucidating the data flow of critical parameters. This enhances and accelerates parameter-tuning efforts. Using a radiation dose calculation program as a case study, this paper demonstrates that the proposed method effectively identifies potential root causes of anomalies, simplifies the direct model, clarifies the interrelations among critical parameters, facilitates parameter tuning, and reduces risks associated with human error.