Market Risk Identification of Power Generation Companies Based on Improved SVM

With the deepening of market-oriented reform of electric power, credit risks of market players occur frequently, seriously undermining the fair and healthy trading environment of the electric power market. Market power risk presents the form of multiple, difficult to regulate, the impact of the characteristics of the big, increasing the difficulty of the construction of the power market. At present, the domestic market regulatory system is not sound, risk identification methods lag behind, market violations occur from time to time. How to efficiently identify the irregularities generated by the market force is of great significance to steadily promote the market-oriented reform of electricity. To this end, this paper combines the index system, which is deeply characteristic of Chinese economy, with the support vector machine, which can realize automatic identification, and proposes an improved support vector machine power generation enterprise violation identification method, which realizes real-time access to the list of companies suspected of abusing the market force, and reduces the workload of the expert judgment to the greatest extent. Finally, the test is carried out using power market simulation data, and the results show that the identification method can identify the abuse of market power behavior of power generation companies quickly and effectively.