Power grid attack detection method and device based on improved RNN neural network

The invention provides a power grid attack detection method and device based on an improved RNN neural network. Various related data in the intelligent power grid are obtained, processed and input into a neural network model for training, an improved genetic algorithm and a particle swarm optimization algorithm are adopted in the neural network model to process the obtained data, and an optimal weight and an optimal threshold value are obtained; wherein the training of the neural network model adopts error back propagation to update a weight and a threshold value, loop iteration is carried outby calculating the mean square sum of network errors as a condition to complete network training, and a detection result is obtained according to data in an intelligent energy grid acquired in real time. According to the method, the genetic algorithm is improved by adopting the particle swarm optimization algorithm, and the data is evaluated by adopting the adaptive function, so that the situation of local concentration