New energy power system elastic optimization method based on deep reinforcement learning

Abstract

The invention provides a new energy power system elastic optimization method based on deep reinforcement learning, which can greatly reduce the labor cost generated in the scheduling process and avoid the influence caused by the time delay of manual communication. According to the method, deep reinforcement learning of the neural network is introduced, learning can be carried out in a highly dynamic random environment through trial and error and exploration, the interactive situation enables the method to have high learning ability and adaptive ability, and the method is more suitable for solving the problem of active power scheduling of a new energy power system with complex nonlinearity and uncertainty. The deep reinforcement learning has both the decision-making ability of reinforcement learning and the computing ability of deep learning, and the application of the deep reinforcement learning in the new energy power system will change the traditional energy utilization mode, so that the system is more intelligent.

Description

technical field [0001] The invention relates to the technical field of power system elasticity, in particular to a new energy power system elasticity optimization method based on deep reinforcement learning. Background technique [0002] In recent years, as the penetration rate of new energy in the power grid has increased year by year, by the end of 2020, my country's installed capacity and power generation of renewable energy will rank first in the world. When a large number of new energy sources are connected to the power grid, they also bring many problems. Due to the characteristics of strong randomness, obvious intermittency, and large fluctuation range of new energy sources, this increases the difficulty of active power dispatching, peak regulation, and frequency regulation of the system, as well as the risk of stable operation. . At the same time, as the largest and most complex man-made dynamic system in the world, the power system is extremely vulnerable to extrem...

AI Technical Summary

Problems solved by technology

Due to a certain delay in manual communication, it cannot meet the intelligence and flexibility requirements of the new energy grid

Images