Health state evaluation and adaptive fault diagnosis decision of key equipment in an integrated energy system based on multi-source heterogeneous data

The data sources of an integrated energy system are complex, including sensors, controllers, computers, and other devices. The data quality may be affected by noise, missing, and outliers, affecting the accuracy and effectiveness of fault diagnosis. Therefore, a new multi-source heterogeneous data-based health state assessment and adaptive fault diagnosis decision of key equipment in integrated energy systems is proposed. Firstly, multi-source heterogeneous data is denoised. Based on this, the LightGBM decision tree evaluation model is constructed to evaluate the device’s health status. The local mean decomposition method is introduced to realize adaptive fault diagnosis of key equipment in an integrated energy system. The experimental results show that the research method can accurately detect the equipment fault, and the application effect is good.