Yang Jing, Wen Zhongjing, Song Qiang, et al. Collaborative Evaluation Method for Terminal and Electric Meter Health States Based on Heterogeneous Data Feature Fusion[J]. 2026, 29(7): 13--.
Yang Jing, Wen Zhongjing, Song Qiang, et al. Collaborative Evaluation Method for Terminal and Electric Meter Health States Based on Heterogeneous Data Feature Fusion[J]. 2026, 29(7): 13--.DOI:
Collaborative Evaluation Method for Terminal and Electric Meter Health States Based on Heterogeneous Data Feature Fusion
Due to the system isolation between terminal devices and electric meters at the data acquisition level
the correlation information between their health states is missing
leading to one-sided evaluation results and reduced accuracy. To address this
a collaborative health state assessment method based on heterogeneous data feature fusion is proposed. By constructing a deep feature fusion network for heterogeneous data
the operational characteristics of terminal devices and the measurement characteristics of electric meters are effectively integrated. An interpretable health state evolution rule library is established
and innovative approaches such as variable-weight theory and graph neural networks are introduced to build a dynamic collaborative assessment model
enabling joint and precise evaluation of the health states of both types of devices. Experimental results demonstrate that the evaluation score deviation between terminal devices and electric meters is less than 4.4%
indicating high accuracy in the collaborative health state assessment of these devices.