YE Bing

School of Science,

Hefei University of Technology,

Hefei 230009, China

FEI Yetai

School of Instrumentation,

Hefei University of Technology,

Hefei 230009, China

LIAO Benqiang

School of Science,

Hefei University of Technology,

Hefei 230009, China

Abstract: The dominant and recessive effect made by exceptional interferer is analyzed in measurement system based on responsive character, and the gross error model of fuzzy clustering based on fuzzy relation and fuzzy equipollence relation is built. The concept and calculate formula of fuzzy eccentricity are defined to deduce the evaluation rule and function of gross error, on the base of them, a fuzzy clustering method of separating and discriminating the gross error is found. Utilized in the dynamic circular division measurement system, the method can identify and eliminate gross error in measured data, and reduce measured data dispersity. Experimental results indicate that the use of the method and model enables repetitive precision of the system to improve 80% higher than the foregoing system, to reach 3.5 s, and angle measurement error is less than 7 s.