Investment Risk Quantification Of Provincial Power Grid Adapting To The Differences In Development Paths

Affected by factors such as spatial-temporal differences,inaccurate risk identification,and differences in personal preferences,various risks overlap and fluctuate in the process of power grid investment,which directly affects the accuracy of power grid investment decisions.Firstly,risk factors are diversified,which are continuously transformed in space and time.Secondly,risk identification is subjective and lacks the discrimination of risk coupling.Thirdly,risk quantification is disordered,it’s difficult to simplify risk information and solve the curse of dimensionality in the process of risk quantification.In order to address these issues,this paper conducts research from three aspects: risk identification,risk decoupling and risk quantification,and establishes risk identification and risk quantification models of the provincial power grid investment adapting to the differences in development paths.(1)Aiming at the problem that risk fluctuation in space and time dimensions leads to low accuracy of risk quantification,investment risk identification models under different development paths of the provincial power grid are established to identify the time-scale characteristics of risks.Firstly,uncertain factors existing in the three dimensions of the development path of the power grid are identified,and the fuzzy theory is adopted to identify key uncertain factors,which realizes the simplification and identification of key uncertain factors.Then,risk identification models considering the characteristics of risk factors are established.The radar map based on equal intercept transformation is adopted to identify the policy risks with sudden change,and the GA-BP neural network model is presented to identify the continuous characteristic of economic development and power grid development risks.Based on the probability distribution fitting and nonlinear regression fitting of the risk,a simplified analysis method of the risk on different time scales is proposed,to determine the distribution regularity of the risk on different time scales.(2)Aiming at the problem that the overlap of risk information leads to the low accuracy of multiple risk quantification,a method for judging the coupling of various risks and a method for decoupling and dimensionality reduction are presented.The correlation analysis method is adopted to judge the inde pendence and coupling of each two risks,and the concept of risk group is put forward,the interpretative structural modeling based on the correlation analysis method is presented,to judge the independence and coupling of risk group.A dimensionality reduction method combining the principal component analysis and the interpretative structural modeling is proposed,which centered on key risk factors to degrade the coupling relationship at the upper layer.In the determination of key risks,the principal component analysis is adopted to extract the most relevant risk factors for solving the missing problem of the risk model.(3)Aiming at the problem that the full-time-scale fitting is difficult to achieve due to the fluctuation of risk on the time scale,an investment risk simplification method and a level quantification method adapting to spatial-temporal differences are presented.A risk quantification method based on the interval number is proposed,combined with the time scale partition method,to simplify the risks with typical distribution characteristics;combined with the risk matrix method,the timeless regularity risk is simplified,to solve the computational disaster problem caused by the variation and fluctuation of risks in the time domain.A comprehensive quantification method of multiple risks considering the characteristics of risk probability distribution is presented,which models the stochastic process of the risk and performs the comprehensive and probabilistic calculation of large-dimensional risk elements.The expected value of risk probability sequence is calculated in consideration of uncertainty of the risk,and the risk level of provincial power grid investment risk is determined.In summary,considering the spatial-temporal differences,uncertainty and coupling of risks,risk identification and risk quantification models of the provincial power grid investment adapting to the differences in devel opment paths are established.The case studies show that the models improve the accuracy of the quantification of single risk and multiple risks in power grid investment,and are beneficial for investors to make reasonable investment decisions.