“Source-load” multiple timing uncertainties and the steady-state/dynamic diversity of reactive power equipment in a new type of distribution system (NDS) determine the need for joint optimization of reactive power allocation and operational regulation. Therefore, based on the “planning-operation” fusion framework, research on optimal allocation of reactive power resources in the “source-load” multi timing joint scenario of NDS is carried out. First, based on the development trend of new forms of NDS, a “planning-operation” joint optimization framework integrating big data analysis, multi time domain operation simulation, and comprehensive benefit evaluation is constructed. Then, based on capacitor steady-state and DSTATCOM dynamic reactive power equipment operation characteristics analysis, sequential optimization of location and capacity is determined. A “source-load” multi-scale joint timing scenario is constructed based on the K-means clustering method, and simulation analysis of NDS operation is achieved in hierarchical reactive power balance through coordination of steady-state/dynamic reactive power devices and deep mining of multi-scale adjustment resources. An NDS reactive power “planning-operation” evaluation index system is established to conduct comprehensive benefit evaluation of the entire process. Finally, taking the IEEE33-bus distribution network system as an example, a reactive power “planning-operation” joint optimization simulation analysis is conducted. The results show that the proposed model has the ability to adapt to the "source-load" uncertainty and multiple timing joint scenarios, and the planning scheme with embedded multi time domain operation simulation is more comprehensive and practical.