Agent-based Simulation Of China's Technological Innovation And Regional Rvolution Of Industrial Structure

In recent years,China's industry is facing pressure from low-end global value chain locks led by Western developed countries.In order to achieve the innovation and upgrade of the industrial structure,the report of the 18 th National Congress of the Communist Party of China in 2012 proposed the implementation of “innovation-dr ive n development strategy” and technological innovation as The important driving force to promote the evolution of industrial structure has affirmed the key role of technologic a l innovation in the process of industrial structure transformation and upgrading and the rise of emerging industries.Based on the previous research,this paper uses the interregional input-output table to construct the economic links between various provinces and departments in China,and uses evolutionary economics theory to simulate the technological innovation of regional sector industries,including product innovation and process innovation,using the subject.The simulation method performs microscopic simulation of industria l structure evolution and presents the simulation results in a macro-emerging manner.In addition,in order to simulate the limited rational behavior of the department's investment in innovation,the DQN model in deep reinforcement learning is introduced into the economic system model to make the behavior of the self-agent more realistic.After verifying the authenticity of the simulation results and the effectiveness of the DQN model for innovative competition,two scenarios were set up for simulation,and the evolution of the industrial structure and production technology level was analyzed.The main conclusions of this paper are as follows:(1)Deep reinforcement learning can guide the limited rational behavior of the self-subject in the proportion of innovation investment alocation,and create a fierce innovation competitio n environment.In the interaction of various departments,there will be a relative ly advantageous department appear.(2)The proportion of primary industries in all provinces is in a downward trend,and provinces with innovative competitive advantages also rely more on fixed capital investment to create higher production scales.The production technology level of primary industries has risen in two scenarios The range is relatively limited.(3)The proportion of the secondary industry in the whole country shows an aggregation phenomenon.The change in the proportion of the secondary industry is mainly caused by the change in the proportion of the manufacturing industry,while the change in the production technology level of the secondary industry mainly comes from the mining industry and the manufactur ing industry.Changes in manufacturing;manufacturing is the main source of advantages for provinces that are technologically innovative and competitive in the secondary industry.Poor performance in the construction industry is the main reason for the province's competitive disadvantage in innovation.(4)The proportion of China's tertiary industry in the northern and southeastern coastal areas has increased significantly,while the provinces with the largest increase in the level of tertiary industry production technology have a certain overlap with the distribution of the Sshaped curve in densely populated areas of China.Industry has played an important role in the innovation and competitive advantages of the tertiary industry production technology level.(5)For industries and sectors that develop faster,we should pay attention to the role of production technology in development,try to avoid pure production expansion and achieve industrial technology upgrades;for industries and sectors with innovative competitiveness,we should Strengthen the distribution and management of innovation investment and give full play to its potential innovation potential.