The invention discloses a day-ahead planning mode reactive power optimization method adaptive to power grid load change, and belongs to the technical field of electric power systems and automation thereof. According to the method, an independent static reactive power optimization model of a plurality of planning modes is solved through the optimal power flow, and the optimal compensation amount ofeach capacitive reactance device node in each mode is obtained. The day-ahead planning mode is subjected to time period division according to the load change situation and the load change threshold value of different voltage partitions, and the capacitive reactance device node shift with minimal segmented total network loss is solved based on linear programming for different operation modes in the same segment. The number of sections is continuously reduced, and iteration is repeated until all the capacitive reactance device meet the action frequency requirement, the segmented optimal shift of the capacitive reactance device node is fixed, and the power generator reactive power and other continuous solutions of a plurality of day-ahead planning modes are further corrected through the optimal power flow, and finally an optimization mode meeting various constraints is obtained. A day-ahead optimization reactive plan being adaptive to the load change condition of a power grid and meetingthe action frequency requirement of a capacitive reactance device can be provided.