Study on Roadway Fault Diagnosis of the Mine Ventilation System Based on Improved SVM

The abstract discusses the importance of reliable mine monitoring systems for preventing accidents. It focuses on diagnosing ventilation system failures caused by roadway faults that alter airflow resistance. The paper proposes using an improved Support Vector Machine (SVM) method to identify the fault location based on changes in air volume detected by wind speed sensors. The approach involves building a sensitivity 0-1 matrix and establishing a roadway fault scope library. The improved SVM method is tested on the Daming coal mine, demonstrating its effectiveness in diagnosing fault locations, reducing training time, and improving accuracy compared to traditional SVM.