Optimal power flow rapid solving method combining neural network and physical model

The invention discloses an optimal power flow rapid solving method combining a neural network and a physical model, and the method comprises the following steps: obtaining a training set based on to-be-input data; based on the neural network and the physical model, constructing an optimal power flow solving model; training an optimal power flow solving model based on the training set; acquiring an under-learning sample, and identifying the under-learning sample based on the trained optimal power flow solving model; and merging the identified under-learning sample into the training set, and circulating the steps until no new under-learning sample is identified, thereby realizing high generalization and rapid solution of the optimal power flow solution model. According to the optimal power flow rapid solving method in combination with the neural network and the physical model, the generalization ability of the trained neural network to the input set can be remarkably improved, and a high-reliability result can