Abstract: Regional innovation is the foundation of national innovation, and it is the growth point and power source for building an innovative country. The goal of a regional innovation policy is to enhance the competitiveness of the regional economy, which must be achieved by strengthening regional innovation capabilities. At present, more and more innovation policy reformers have proposed that there is no “one-size-fits-all” innovation policy applicable to all types of regions. The perfect regional innovation policy design needs to take into account the innovation foundation, innovation efficiency and innovation potential of different regions, base on the regional innovation situation, and make differentiated flexible adjustment based on the regional innovation pattern. This study intends to divide China′s regional innovation pattern by referring to the knowledge differentiation model theory and the "smart innovation policy" system, in order to form regional innovation policy system based on knowledge growth, creating" knowledge base "characteristics in different regions of diversified innovation growth pole, it is of great theoretical value and practical significance for the top-level design of China′s regional innovation strategy.Firstly, this paper reviews in detail the knowledge differentiation model of regional innovation pioneered by European scholars. The KIT project in Europe divided the innovation patterns of 262 EU regions through empirical methods. Through cluster analysis of measurement indicators, the research team found five categories of innovation pattern: science-based area (pattern 1), applied science area (pattern 2),smart technological application area (pattern 3), smart and creative diversification area (pattern 4) and imitative innovation area (pattern 5). Based on the empirical classification, the "Smart Innovation Policy" was formally proposed by Camagni and Capello. Two scholars define this policy as a dynamic policy that can able to increase the innovation capability of an area and to enhance local expertise in knowledge production and use, acting on local specificities and on the characteristics, strengths, and weaknesses of already-established innovation patterns in each region. It focuses on R&D elements, and emphasizes that policy design must consider the two principles of "embeddedness" and "connectedness", that is, innovation policies have to be embedded in the local reality, in local assets and strategic design capabilities, and have to guarantee the achievement of external knowledge through strong and virtuous linkages with the external world. The purpose of this policy is to design targeted support measures for each type of "regional innovation pattern" through external policy interventions to enhance the effective characteristics of the innovation pattern and enhance innovation efficiency. Camagni and Capello focus on regional innovation patterns in terms of policy goals, policy actions for local knowledge generation (embeddedness), policy actions for exploitation of knowledge spillovers (connectedness), policy style, and policy beneficiaries. According to the innovation pattern of different regions to carry out the differential design, forming a complete set of policy guidelines.Secondly, referring to the European regional innovation pattern and measurement index system, the quantitative evaluation of the innovation factors and knowledge base of 31 provinces, cities in China was conducted. Specifically, five first-level indicators were designed: knowledge creation, innovation creation, regional preconditions for knowledge creation, regional preconditions for innovation creation, Inter-regional knowledge and innovation flows. Subsequently, 11 secondary indicators were designed under the primary indicators. Through cluster analysis, this paper divides into five types of regional innovation patterns, which are science-based area, applied science area, creative application area, imitative innovation area, and innovation enlightenment are. In this paper, ANOVA was used to analyze the innovation factor indexes of five types of patterns. The mean values of each index were calculated and the significant differences were tested. Combining the representative regions of each pattern, it can be seen that China′s regional innovation elements and innovation capabilities show obvious characteristics of strong east and weak west. In addition, due to the obvious differences in industry foundation and region, the five regional innovation patterns in China have different characteristics.Finally, this paper makes a differentiated redesign of China′s regional innovation policy system according to the situation of innovation knowledge in differentregions.In view of the five types of regional innovation patterns in China, this paper proposes the following suggestions: (1) adopt the "bottom-up" mode and encourage the differentiation and contextualization of innovation policy objective design; (2) the science-based area (pattern 1) is in the leading position in innovation capacity and resources in the country, and the goals of innovation policies should clearly maximize the return on R&D investment. We should make full use of the advantages of local scientific research and technical talents, integrate existing intellectual resources, establish a stable network of industry-university-research cooperation, and encourage research institutions, universities and enterprises to actively participate in the design and decision-making of regional innovation policies; (3) the innovation policy objectives of applied science area (pattern 2) and creative application area (pattern 3) should be clear to maximize and specialize the return of application cooperation, and quickly transform the innovation of basic scientific research into commercialization and industrialization results. It is important to establish a cross-regional coordination mechanism, optimize the spatial distribution of innovation resources, concentrate innovation resources, and support superior industries to become more sophisticated and stronger; (4) the innovation policy goal of the imitative innovation area (pattern 4) should be positioned to maximize return to imitation and stimulate the vitality of enterprises with the innovation of system and mechanism as the core. Based on the existing foundation of innovation, we need intensively cultivate in the advantageous and characteristic industries, build an innovation platform for the government, scientific research institutes and enterprises to share and promote the transformation of government functions; (5) the innovation policy goal of the innovation enlightenment area (pattern 5) should position and focus on industries with ecological characteristics and build an endogenous innovation drive based on the application of green technologies. It is necessary to combine the actual development of the region, develop distinctive products and brands, provide corresponding technical guidance and financial support for the advantages of local innovation resources, promote the rational flow of talents and other scientific and technological resources, enhance the competitiveness and entrepreneurship of local enterprises, improve the overall innovation capacity of the region.and promote the regional innovation pattern to evolve into applied science area or creative application area.

Key words: knowledge differentiation model, smart regional innovation policy, policy practice