On fault prediction based on industrial big data

Fault-free production is a basic characteristic of intelligent workshop. The existing model-based fault prediction methods depend much more on the precise models of the equipment, while the data-driven ones can only use some basic state data with very limited volume. These features make them not only impractical, but also not able to meet the real-time requirement of analyzing industrial big data under the environment of Industrial Internet of Things. This paper presents a fault prediction method based on industrial big data, which directly excavates the relationship between the data such as the sound and status data, and the equipment faults by machine learning methods. What is more, the equipment state can be monitored in real time leading to the failure would be checked out timely. The simulation result shows that our method has high accuracy and real-time features compared with the existing ones.