Data-Driven Intelligent Fault Diagnosis Technology for Transmission Lines of Wind Power Renewable Energy System

The vigorous development of wind power is of great significance for constructing the new type of power system. In the wind power renewable energy system, AC line faults exhibit characteristics different from traditional faults under the conditions where phase angles are controlled and amplitudes are limited. This may lead to false trips or refusals by traditional protection methods. In order to solve this problem, this paper investigates the line fault characteristics of the wind power renewable energy system in double-ended weak feed systems, then selects statistical measures that reflect fault characteristics in both the time and frequency domains. Finally, this paper proposes an intelligent diagnosis model for line faults in wind power renewable energy systems based on SVM (Support Vector Machine). A simulation model incorporating wind power was established using the PSCAD software. By adjusting the fault resistance, fault location, and fault type, data for the three-phase voltage and current of the AC side fault line are collected, basic and advanced electrical characteristics of the line in both time and frequency domains are obtained. 80% of this data is selected as the training set, and the fault characteristics are trained based on the support vector machine algorithm. The remaining 20% is used as test data. The test results show that the fault identification accuracy reached 99.68% and the protection scheme can accurately and quickly identify faults.