A Causal-Guided Multimodal Large Language Model for Generalized Power System Time-Series Data Analytics

TL;DR

This paper introduces a causal-guided multimodal large language model for power system time-series analysis, capable of handling diverse tasks with improved accuracy and efficiency through causal discovery, multimodal encoding, and adaptive input-output mechanisms.

Contribution

It presents a novel CM-LLM framework integrating causal discovery, multimodal data fusion, and adaptive task handling for heterogeneous power system time-series analysis.

Findings

Achieves high accuracy on missing data imputation, forecasting, and super-resolution tasks.

Demonstrates effectiveness with simple fine-tuning on real-world datasets.

Enhances model flexibility and resource efficiency in power system analytics.

Abstract

Power system time series analytics is critical in understanding the system operation conditions and predicting the future trends. Despite the wide adoption of Artificial Intelligence (AI) tools, many AI-based time series analytical models suffer from task-specificity (i.e. one model for one task) and structural rigidity (i.e. the input-output format is fixed), leading to limited model performances and resource wastes. In this paper, we propose a Causal-Guided Multimodal Large Language Model (CM-LLM) that can solve heterogeneous power system time-series analysis tasks. First, we introduce a physics-statistics combined causal discovery mechanism to capture the causal relationship, which is represented by graph, among power system variables. Second, we propose a multimodal data preprocessing framework that can encode and fuse text, graph and time series to enhance the model performance. Last, we formulate a generic "mask-and-reconstruct" paradigm and design a dynamic input-output padding mechanism to enable CM-LLM adaptive to heterogeneous time-series analysis tasks with varying sample lengths. Simulation results based on open-source LLM Qwen and real-world dataset demonstrate that, after simple fine-tuning, the proposed CM-LLM can achieve satisfying accuracy and efficiency on three heterogeneous time-series analytics tasks: missing data imputation, forecasting and super resolution.