The efficiency and numerical stability of power system dynamic simulation are often affected by the rigidity of studied systems. It is indicated that rigidity has space-time local characteristic that offers the foundation to solve the rigidity problems by means of the analysis and re-thinking about the rigidity characteristic of dynamic processing of nonlinear system. Secondly

self-adaptive flexible grouping strategies and groups coordination strategies with flexible self-adjustment method are put forward. By unified treatments of three adaptive strategies including variables grouping

relative step factor adjustment and minimal integral step size adjustment

changing of groups and step sizes can be controlled dynamically. They magnify the step sizes of some slower variables under the allowable precision

which avoid using the same step size for all of the variables as the traditional method

and overcome the defect of the present grouping methods that are difficult to adjust adaptively multi control parameters at the same time

and then the efficiency of dynamic simulation is enhanced significantly.