A real time flight deck safety monitoring system based on support vector machine

In complex system, the safety incidents and accidents often result in a great loss of personnel and equipment. However, the traditional alarming system and applications might not capable to meet the requirements of system safety control. In engineering applications, the lack of accidents samples impacts the accuracy of predictions for potential accidents, how to use a small amount of observational data to assess the relationship between operation data and safety has become an important issue in prediction and assessment of system safety. Considering the lacking of incident samples during system operations, we purposed a system safety monitoring and trend predicting method based on support vector machine (SVM), established a system safety trend prediction model and processes, the case application verified the validity and accuracy of the method.