An SVR-Based Online Fault Detection Method

Many complex engineering systems have the characteristics such as difficult modeling, dangerous testing, and high cost to experiment with fault. Aim at these points, a new framework for online fault detection is presented. This framework includes a mechanism for associating each detection result with a confidence value. Based on this framework, a concrete real-time fault detection method is developed by an online support vector regression algorithm. The method can detect predict fault quickly without any data and prior knowledge of fault, so it is provided with strong practical significance. The simulation on the fighter F-16 is performed to demonstrate the promising performance of the method.