Aiming at the potential risks existing in the power market transaction under the new power system, and considering the temporal attributes of the information, this paper proposes to use dynamic Bayesian network to construct the risk monitoring and early warning model of the power market transaction. The dynamic Bayesian network is utilized to calculate the correlation between different risk factors, estimate the risk value of power market transactions, and classify the warning level. Taking the southern regional electricity market as the research object, the relationship between electricity price and transaction volume is explored based on the experimental dataset. A credit grading system is introduced to carry out transaction prediction simulation experiments, relying on the prediction data to determine the link between electricity price and transaction volume. The results show that overall power price and transaction volume show a negative correlation, but in June, when the power price is 0.4370 yuan per kWh, the transaction volume still reaches 19.65 million kWh, and the inverse relationship between the transaction volume and the price is not obvious. The use of dynamic Bayesian network to construct the power market transaction risk monitoring and early warning model can dynamically update and adjust the risk monitoring with the passage of time, making the power market transaction early warning more flexible and real-time.
1970-2025 CP (Manitoba, Canada) unless otherwise stated.