Research on Multi-Agent Requirements Uncertainty of Complex Product Based on S-Cloud Model

The uncertainty of customer requirements itself is a major problem in the analysis of complex product requirements. Sentiment analysis is very important in the analysis of requirements uncertainty. However, the current researches ignore the different degrees of ambiguity of requirement agents in different positions in sentiment expression. A multi-agent requirements mining and analysis method for complex products based on S(sentiment)-cloud model considering uncertainty is proposed. Firstly, based on the multi-agent requirements mining model, the requirements of different requirement agents are obtained from different data sources. The relative weight between the requirement agents of complex products is determined by the power-interest matrix. Secondly, a requirement analysis method of S-cloud model is proposed, which quantifies the sentiment tendency of customers and transforms it into a multi-level evaluation S-cloud model. The S-cloud model is adjusted according to different requirement agents, the multi-agent requirements weight is obtained by fusion. Finally, a prototype system for multi-agent requirements mining and analysis of complex products is developed. Then the feasibility of this method is verified by taking high-speed trains as an example.