Fault Diagnosis of Integrated Energy System Based on CNN-LSTM

The integrated energy system (IES) has the characteristics of electricity, heating, cooling and gas multi-energy coupling. Compared with the single energy system, its fault characteristics and diagnosis are more complex, and the fault diagnosis of the integrated energy system has become particularly important. To solve this problem, this paper proposes a fault diagnosis method of integrated energy system based on convolutional neural network (CNN) and long short memory network (LSTM). First of all, use the built park integrated energy system model to generate normal samples and fault samples of typical equipment of the integrated energy system. Establish the RT-LAB integrated energy system park model based on digital twinning, and then use the model to generate normal samples and fault samples of typical equipment in the integrated energy system. Secondly, select the fault eigenvalue index and extract the fault characteristics of the typical equipment in the integrated energy system, use the principal component analysis method to reduce the dimension of the features, and then use the CNN-LSTM neural network model for training, to achieve the purpose of accurate identification of combined heating and power (CHP) system fault in the integrated energy system. The result of an example shows that the accuracy of the proposed method is 95.69%, which can effectively solve the problem of fault diagnosis of integrated energy system.