NON-LINEAR FUZZY RECOGNITION AND ITS APPLICATION IN IDENTIFYING SST ABNORMALITY

Astract: Based on the principles of non linear fuzzy inference and recognition,a method and approach identifying systemic chief influence factors from an actual signals and accordingly removing its interference,was studied and discussed, a relevant denoise experiments was also carried out.From our experimental results, it is shown that due to the affiliated advantages of fuzzy inference system such as non linear, tolerance error and self adaptable, by fuzzy reasoning, we can easily identify and recognise the chief influence factors from a non linear complicated system and effectively find out its contribution to the system.As an application, based on appointed observational data, the research process of identifying chief influence factors forcing on El Nio/La Nia was explored , and the chief inducing/exciting effect of the pacific trade wind influencing on El Nio/La Nia events since 1970's were diagnosed and identified. It is shown that the notable El Nio events during 1970's were mainly induced by the west pacific west wind near the equatorial being abnormally stronger, on the other hand, the remarkable El Nio events during 1980's (especially in the 1982-1983's SST warming event) were chiefly leaded by the corporate effects of the equatorial trade wind abnormality both over the west and the east Pacific, generally, first exciting by the former, then strengthening by the latter.