A reactive power control method for power grid based on deep generation load forecasting

Abstract

The invention relates to a power grid reactive power control method based on deep power generation load forecasting. Firstly, the deep learning method is used to make time-series progressive and accurate predictions of day-ahead and real-time power generation and load; then, according to the predicted value, the idea of ​​reinforcement learning is used to optimize the adjustment of time-series power flow. The invention realizes the intelligent balance of power flow and reactive power of the power grid through the precise prediction of new energy power generation and load and the automatic adjustment of reactive power of the power grid by means of reinforcement learning.

Description

technical field [0001] The invention relates to the field of grid planning and protection operation, and more particularly, to a grid reactive power control method based on deep power generation load prediction. Background technique [0002] There are many types of traditional modern power grid loads and distributed renewable energy sources, and the output fluctuates greatly. The output changes of different types of loads and distributed power sources cause deviations between the node reactive power data of grid power flow calculation and the actual node reactive power, which affects the accuracy of power flow calculation. The existing AVC is an ex post control strategy, which cannot effectively deal with the reactive power regulation requirements of the grid under the situation of new energy generation and load reactive power fluctuations, and frequent regulation will affect the service life of the regulating equipment. In order to solve this problem, the present invention...

AI Technical Summary

Problems solved by technology

[0002] There are many types of traditional modern power grid loads and distributed renewable energy sources, and the output fluctuates greatly

Different types of loads and changes in the output of distributed power sources cause deviations between the reactive power data of the grid power flow calculations and the actual node reactive power, which affects the accuracy of power flow calculations

The existing AVC is an ex-post control strategy, which cannot effectively meet the reactive power adjustment requirements of the grid under the circumstances of new energy generation and load reactive power fluctuations. Frequent adjustments will affect the service life of the adjustment equipment.

Images