Hybrid artificial intelligence for power grid line fault diagnosis and restoration auxiliary decision-making

This paper proposes a data-driven power grid line fault diagnosis and knowledge-driven auxiliary decision-making technology for intelligent restoration of power lines. First of all, a confidence-expert system fault diagnosis model for power grid lines is proposed, and the model misjudgment is avoided by using the confidence model to score the evaluation results of the expert system. Then, the ontology structure of the knowledge atlas for power grid restoration decision is designed, and the power system named entity recognition model is trained to assist the automatic generation of the knowledge atlas. Finally, the active retrieval technology of restoration knowledge based on fault diagnosis results of power lines is constructed to realize intelligent research and judgment of restoration information of fault lines. According to the experimental data, the accuracy of the newly proposed line fault diagnosis technology is about 97.6%, which is about 20% higher than that of the traditional expert system method. The retrieval and judgment time of the fault line power recovery disposal information is less than 1s, which greatly improves the recovery efficiency and ability of the power system after the line fault, and reduces the workload of the dispatchers to collect the operation information and formulate the power recovery strategy when dealing with the fault.