Electricity market power generator market power monitoring method based on random forest regression

Abstract

The invention discloses an electricity market power generator market power monitoring method based on random forest regression. Generator characteristic data and label data in a certain historical period of time are selected as training samples, the training samples are input into a random forest regression algorithm toolkit, default parameters in the toolkit are used as algorithm parameters, training fitting is performed, and a generator market force monitoring model based on random forest regression is obtained through output; when a target power generator is monitored, the prediction sampleinput value is input into a power generator market force monitoring model based on random forest regression, target power generator prediction label data is obtained through output, and if the prediction label data is larger than a set threshold value, it is judged that the power generator carries out market force; and if the predicted label data is smaller than or equal to the set threshold, itis judged that the power generator does not have the market force. According to the invention, the market monitoring efficiency is improved, the effective range of market monitoring is expanded, goodcompetitiveness of the market is ensured, and the market operation efficiency is improved.

Description

technical field [0001] The invention relates to a random forest regression-based method for monitoring the market power of generators in an electric power market, belonging to the technical field of electric power market supervision. Background technique [0002] At present, as the construction of the electricity spot market enters the stage of acceleration, the relevant functional specifications and construction outline of the spot market will be organized and tried out in the pilot areas, and relevant national standards will be issued when the conditions are ripe. [0003] The existing power spot market trial operation plan solves the substantive problems that plague the reform of the power market, such as planning and market, spot transactions and medium and long-term transactions, and competition of units with different costs on the same platform. [0004] The power industry has a natural monopoly. The introduction of market mechanisms in the power market does not mean t...

AI Technical Summary

Problems solved by technology

The behavior test process needs to simulate the clearing calculation of the market, which will consume a lot of time and is not conducive to improving the monitoring efficiency of market power monitoring. If there are no commercial operators, they will escape the monitoring of the behavioral impact test and narrow the effective scope of market power monitoring, which is not conducive to the realization of good competition in the electricity spot market, and cannot better improve market operation efficiency and maximize social welfare, thus highlighting the improvement of the traditional Behavior affects the need to test the model

Images