Application of Multi-sensor Data Fusion Method Based on DS Evidence Theory in Fault Diagnosis

Overview of Multi-Sensor Data Fusion

In the fault diagnosis of turbogenerators, sensors are one of the essential components in turbogenerators, they can feed back various parameters and states to us, multi-sensor data can obtain information about the same object from multiple aspects and To be comprehensively utilized to improve diagnostic efficiency, multi-sensor data fusion can make diagnosis more accurate and reliable. However, due to the limitations of the sensor itself and the differences between different sensors, it is difficult to draw completely accurate conclusions even if we obtain a large amount of data. Therefore, in order to fuse these data together to obtain more reliable diagnostic results, it is necessary to adopt a decision-level fusion method.

In decision-level fusion, we need to evaluate and make decisions on each decision according to certain criteria in order to select the optimal decision. These criteria include the accuracy of sensor information, the uniformity of data distribution, the number of sensors, the complexity of fault types, and so on. After determining these criteria, we need to comprehensively consider the judgment results of each sensor in order to obtain the final diagnosis result.

In practical applications, sensor data fusion is often achieved through models and algorithms. Among them, DS evidence theory is a very commonly used method,