Research on Fault Diagnosis of Converter of Doubly-Fed Wind Power System Based on Deep Learning Algorithm

The converter fault diagnosis is critical in doubly-fed wind power generation system, however it has an issue with erroneous performance positioning. The typical Neural network algorithms is unable to address the inaccurate fault location issue in doubly-fed wind power generation system, and the result is insufficient. As a result, a Deep learning algorithms-based Research on fault diagnosis of converter of doubly-fed wind power system is provided, and tResearch on fault diagnosis of converter of doubly-fed wind power system is assessed. To begin, the gradient descent theory is used to discover the influencing elements, and the indicators are split based on the converter fault diagnosis's needs to decrease interference factors in the converter fault diagnosis. The gradient descent theory is then used to create a Deep learning algorithms converter fault diagnosis scheme, and the outcomes of the converter fault diagnosis are thoroughly examined. The MATLAB simulation results reveal that, under particular evaluation conditions, the Deep learning algorithms outperforms the standard Neural network algorithms in terms of converter fault diagnosis accuracy and time of influencing variables.