A dialogue-driven framework for intelligent and automated simulation of complex power systems

Abstract

Regarding the growing complexity of power grids and the increasing workload for operators, this paper presents a dialogue-driven framework for intelligent and automated simulation of power systems. This framework facilitates cooperative grid regulation by leveraging both human and artificial intelligence (AI) through a natural interaction via a dialogue system. We outline five core components and their respective functions, inputs, and outputs, as well as identify four key techniques that warrant further research. Our implemented prototype system assists operators in line flow control tasks, providing technical guidance and demonstrating feasibility. A user study comparing the usability of our prototype reveals its potential to enhance operators’ time efficiency and offers valuable assistance with high acceptance. Furthermore, the system demonstrates generalizability and could inspire other engineering scenarios involving human-AI collaboration.