Abstract: Narrowing the regional differences of manufacturing intelligentization has become the primary problem to comprehensively and coordinately promote the development of manufacturing intelligentization. Therefore, revealing the regional characteristics and differential convergence rule of China's manufacturing intelligentization systematically is of great significance for formulating scientific, reasonable and targeted manufacturing intelligentization policies, and constructing higher quality, more efficient, fair and sustainable coordinated development pattern of manufacturing intelligentization. Firstly, the index system of manufacturing intelligentization is constructed based on the perspective of technological economy paradigm, and measured by entropy weight method. On this basis, the spatial distribution of manufacturing intelligentization is described by spatial correlation, so as to investigate the regional characteristics. Then, the method of theil index and structural decomposition is used to quantify the regional differences of manufacturing intelligentization and reveal the source of the differences. Finally, the spatial β convergence model is used to test the convergence of manufacturing intelligentization in the whole country and three regions, and then explores the convergence rule. It is found that: (1) the manufacturing intelligentization in China has the characteristics of homogeneous agglomeration and unbalanced regional development. The global spatial autocorrelation indicates that the spatial distribution is not completely random, but a higher (or lower) manufacturing intelligentization area, which is often adjacent to one or more higher (or lower) manufacturing intelligentization areas. It shows that the eastern region years has been in the "high-high" agglomeration area, and the western region has been in the "low-low" agglomeration area for many years; Local spatial autocorrelation shows that the development of manufacturing intelligentization has the dual imbalances of between regions and within one region; the spatiotemporal transition shows that the spatial distribution of manufacturing intelligentization has high stability, and there are linkage effects among areas. (2) The overall difference in manufacturing intelligentization is relatively high but decreasing, and the differences within one region is the main source of overall difference. Among them, the eastern region, the contribution rate of theil index average is the largest, highest followed in the western region and central region to a minimum, that to narrow the differences within one region, especially in the eastern region, is an important driving force for the coordinated development of manufacturing intelligentization. (3) Absolute β convergence exists in the eastern, central and western regions, but not in the whole country, indicating that the differences between regions in manufacturing intelligentization will appear for a long time, but the differences within one region will disappear automatically. After considering the factors of technological innovation investment, human capital and informatization level, conditional β convergence exists in the whole country and the three regions, and conditional β convergence rate of the national, eastern and western region is faster than absolute β convergence rate, indicating that after considering the control variables, the differences between regions and within one region inter-regional in manufacturing intelligence will gradually narrow. Among them, human capital is conducive to narrowing the differences between regions and within the eastern and western regions, and the informatization level is conducive to narrowing the differences between regions and within the central and western regions.

摘要： 缩小制造业智能化区域差异，已成为全面协调推进制造业智能化发展面临的首要问题。本文基于技术经济范式视角构建制造业智能化指标体系。在此基础上，刻画空间分布，量化区域差异，检验收敛机制。研究发现：（1）中国制造业智能化存在同质性集聚和区域发展不平衡特征。（2）区域内差异是制造业智能化差异的主要来源，其中，东部区域贡献率最大，西部次之，中部最小。（3）在考虑了技术创新投入、人力资本、信息化水平因素后，区域间和各区域内部的制造业智能化差异将逐渐缩小，其中，人力资本有利于缩小区域间和东、西部区域内部差异，信息化水平有利于缩小区域间和中、西部区域内部差异。