Research On Visibility Graph Model Based On Correlation Coefficient And Fluctuation Law Of Carbon Price

For analyzing big data,not only feasible computing capabilities are required,but also methods for the high-efficiency analysis of the data are needed.As explored in this article,using visibility graph models to construct data network based on time series is one of the best ways to analyze data.Although the existing visibility graph models can continue the inherent morphological characteristics of time series,some performances have certain limitations,such as the complexity of the algorithm and the anti-noise performance.This paper,based on the existing general visibility graph(VG)and horizontal visibility graph(HVG)models,in the construction of the model,combined with the sliding window and the correlation coefficient,proposed a horizontal visibility graph model based on correlation coefficient(CHVG).Next,this paper generalizes the CHVG model,proposes an n-order horizontal visibility graph model based on correlation coefficient(CIHVGn),respectively gives theorem and the proof process of the degree distribution convergence of the complex network obtained by the mapping of the two models,and verifies the correctness of the theoretical results by numerical simulation.This paper combines the CHVG model with VG and HVG to compare and analyze the models through four common time series(namely white noise,period,fractal,and chaotic time series).The research shows that compared to the VG and HVG models,the CHVG model can extract the complex network features of these four time series more efficiently,and the model has stronger anti-noise performance.On this basis,this paper applies the CHVG model to the carbon price data of the EU carbon market and domestic pilot areas respectively,and analyzes the network characteristics of the data,the dynamic evolution process and the linkage relationship between them.The results show that these carbon markets themselves have similarities and scale invariance,and the past carbon prices can be used to predict future carbon prices.Moreover,there are similarities between the domestic and EU carbon markets.Furthermore,this paper uses Euclidean distance and introduces a matrix of related indexes to analyze the network characteristics of carbon price data in the EU carbon market and domestic pilot carbon markets,and uses the distribution characteristics of related indexes γ to describe the dynamic characteristics of carbon price evolution.Based on the research of domestic and foreign carbon markets and drawn lessons from the EU carbon market,this paper puts forward some policy recommendations on the development of domestic carbon market.