Real-time Fault Detection and Disaster Recovery Response System of Information System Based on Neural Network

The study proposes a system architecture that includes four core modules: data collection, neural network processing, disaster recovery response, and system management. This architecture utilizes deep learning techniques, particularly the combination of convolutional neural networks (CNN) and long short-term memory networks (LSTM), to fully utilize time series data and spatial feature information, achieving real-time monitoring and rapid response to information system failures. The experimental results show that the system achieved a fault detection accuracy of 98.5% in the test set, with an average response time of about 2.3 seconds, a data recovery rate of over 99.65%, and a false alarm rate of below 1.30%. These results validate the effectiveness of real-time fault detection and disaster recovery response systems based on neural networks in improving the stability and security of information systems.