Abstract
In nuclear power plant projects, requirements engineering manages the sheer volume of requirements, typically characterized by descriptive and non-harmonized requirements. Large projects may have tens of thousands to hundreds of thousands of requirements to be managed and fulfilled. There are two main issues impeding requirements analysis; tortuous requirements to be interpreted, and humans' very limited ability to concentrate on a specific task. Therefore, it has been recognized that artificial intelligence (AI) algorithms could have potential to support designers' decision-making in classifying and allocating nuclear power plant requirements.
This paper presents our work on developing an AI-based requirements classifier utilizing natural language processing (NLP) and its integration with the requirements management system. The focus is on the classification of nuclear power industry-specific requirements utilizing deep learning-based NLP. Three classifiers are compared with each other and the corresponding results are presented.
The results include predetermined requirement classes, manually gathered and classified data, comparison of three models and their classification accuracies, microservice system architecture and integration of the established classifier with the requirements management system. As the performance of the requirements classifier and related system has been successfully demonstrated, future AI-specific development and studies are suggested to focus on atomizing multi-class requirements, combining similar requirements into one, checking requirements syntax and utilizing unsupervised learning for clustering. Furthermore, new and advantageous requirement classes and hierarchies are suggested to be developed while also improving current datasets both quantitatively and qualitatively.
Suggestion on the safety classification of spent fuel dry storage in China’s pressurized water reactor nuclear power plant
