AVC Planning Strategy of Power Grid Based on Genetic Algorithm Improved Neural Network Quantile Regression

Abstract Aiming at the security risk of power system instability, this paper proposes a parameter optimization AVC planning strategy based on neural network. Based variety grid operation modes, innovatively uses genetic algorithm to improve network quantile regression control variables obtain state variables. This method can reduce frequency invalid operations, make more stable, and security. In actual case, improved improves accuracy by least 29% compared with multiple linear regression.