| abstract |
The present invention proposes a grid reactive power and voltage regulation method and system based on time series reinforcement learning, including: acquiring the voltage amplitude of the node to be regulated and the number of input capacitors; ; For the selected agent, according to the pre-established Q table for the agent, the voltage amplitude and the state of the capacitor that has been put into the current sequence, the reactive voltage of the grid is regulated by acting on the capacitor. In the present invention, the perception of the state of the power grid by each agent includes the voltage amplitude and the input quantity of the capacitor, and the corresponding agent in the group is selected according to the "input quantity of the capacitor". Reactive power voltage, the method proposed in the invention makes up for the defect of the existing research that the reactive power voltage regulation method fails under the influence of load fluctuation and limited reactive power compensation equipment resources, and can better improve the voltage condition of the load center node. |