Discussion Of Several Similar Modeling Methods And Similarity In Qualitative Phase Portraits

Similarity theory, as an important basis of quality evaluation of simulation sys-tem, has played a huge role when people adapt and change the world. It has been re-searched and applied in wide areas such as engineering, physic experiments, model construction, data mining and so on.Qualitative simulation is a method for qualitative modeling of system. According to qualitative relationships between the variables of the system, it gives all the possi-ble behaviors. Qualitative phase portrait is a geometrical state portrait constructed based on qualitative simulation method,which can express the global behavior of the system. The actual behaviors of the system are often represented by time series ob-scured with noise. The time series is a numerical analysis, while the qualitative phase portrait is a qualitatively expression of the system state. It is necessary to evaluate the quality of qualitative phase portrait and discuss their similarities, which is full of dif-ficulties and needs to make a further study on similarity theory.We firstly discusses similarity theory in several fields, and their advantages and limitations are put forward. It's also useful to the similarity discussion of time series and qualitative phase portrait with reference.Then, smoothness noise reduction method is proposed to improve common ways, making the system characteristics to the fore from the noise, in order to facilitate qualitative processing of time series.Finally, the similarity of time series and qualitative phase portrait is discussed on the trend coarse-grained, and some similar indexes are given, such as saddle point, the source points, destination points, separatrices, etc., and we make a similarity analysis in experiments.