In power systems

many variables appear as time series under different scales

such as voltage/current waveforms in transient analysis or power/load data. Among these variables

the relationship between load and other time series has been widely investigated in applications such as load forecasting or system dispatch. Analysis of relationship between time series can not only provide reference for research on unknown variables

but also can reveal the underlying physical laws of power system operation to some extent. However

in most existing research

the concept of relationship stays as correlation

rather than causality

which leads to the weakness of the physical meaning of relationship

limited information provided by some related variables

and the high complexity of the model. Current research concerning causality in power systems focus mainly on causality between two variables

which is hardly applicable for multiple variables. To this end

an improved fast casual inference algorithm based on multiple causality inference method is proposed in this paper and applied to infer the causality between load and multiple meteorological data series. Case studies based on real data from 2019 to 2022 are conducted to compare with a method based on correlation analysis

which validates the effectiveness of the proposed algorithm.