H-M bearing capacity of a modified suction caisson determined by using load-/displacement-controlled methods

Open-circuit fault diagnosis method for inverters using deep learning and the evidence reasoning rule

“…DNN has a multi-hidden layer structure and can perform feature transformation on large-scale fault voltage and current data, thereby extracting deep features from the data. The data processing of the CNN method [16] is simple, and it divides the three-phase current data values into three dimensions, which can improve the learning ability of the model and further improve the diagnostic accuracy. Therefore, although both CNN and DNN belong to the field of deep learning, the two network structures are different, which leads to different performances for the same problem, and each has advantages and disadvantages.…”

“…Chen et al [15] took the current signal on the DC side as input and designed a diagnostic method based on a Convolutional Neural Network (CNN) to monitor the open-circuit status of an IGBT. Shang et al [16] collected three-phase current signals of an inverter and used a CNN for fault diagnosis. Talha et al [17] introduced a feature extraction system based on the three-dimensional three-phase voltage pattern map and located the fault using the neural network.…”