Although China's economy has rapidly increased these years, there still exists an imbalance between developments in different regions. The income of some rural areas is still significantly below the international average level. In order to seek measures to promote rural residents' income growth, this paper used the panel data model to explore the industrial electricity big data considering its timeliness and accuracy, seeking the correlation between electricity consumption and economic development.The study found that the relationship between net income and electricity consumption is strong and could be depicted properly by the panel data model. The study also found that in most poor provinces in China, the income of rural residents was significantly positively correlated with agricultural development, business development and household electricity consumption, however, it was negatively correlated with industrial development. These features provide guidance for policy implementation by government. In addition, the prediction accuracy of the panel model is better than most of machine learning models, due to consideration of the tendency of macroeconomic variables over time.