Research on Stability Prediction Model of Smart Grid Based on Neural Network
At present,machine learning plays a more and more important role in smart grid stability prediction.In view of the shortcomings of many assumptions and inaccurate prediction in the traditional machine learning model,this paper proposes a smart grid stability prediction model based on neural network.The model uses the damped least square method to consider the reaction time to train the data,which takes consumed and produced power and elasticity coefficient as the input variables,to predict the network stability as the output variable.The activation function of the hidden layer is tansig,and the activation function of the output layer is purelin.In this paper,mean square error(MSE)and R-square(R2)are used to evaluate the ac-curacy and effectiveness of the prediction model.The prediction results show that the prediction model has enough accurate pre-diction performance in the training and testing stages,which has very low MSE value and maximum R2 value in the prediction range.
万方数据
维普
