Data-driven power grid analysis and optimization have attracted broad attention in recent years. Firstly

a comparison of thinking modes between data-driven and physical model based methods in power grid analysis and optimization is conducted and the contacts and differences between data-driven and model based methods are put forward. Furthermore

the study subjects of power grid analysis and optimization are classified

the research programs of existing data-driven power grid analysis and optimization are summarized

which are based on the aspects of topology identification

parameter-topology joint identification

system matrix identification

power flow calculation

and optimal power flow calculation and so on. Key technologies of data-driven power grid analysis and optimization are summarized. Finally

challenges and future research trends in this field are proposed and the future research directions are prospected.