A three-stage architecture constructed by combining independent component analysis (ICA), gray correlation coefficient analysis (GCCA) and grey model GM(1, N) is proposed for evolution region innovation. In the first stage, ICA is used as feature extraction. In the second stage, GCCA is used as main factors select. In the third stage, GCCA is used to evolution innovation ability. The paper applies the ICA, grey relational analysis and grey non-linear regression analysis to study the factors affecting the ability of independent innovation in Chinese industrial enterprises from the empirical perspective. It is shown that the proposed method achieves is effective and feasible. And it provides a better estimate tool for the innovation activity. It also provides a novel way for the evolution design of the other engineering.