Anomaly detection in analog circuits using Support Vector Data Description

With the rapid development of electronic technology, electronic systems are widely used in various fields, and the reliability requirement for electronic systems is also increasing. Many test and diagnosis methods have been proposed. However, few methods can be used to recognize the intermediate anomaly state between normal and fault state. To solve the above problem, this paper presents an anomaly detection method based on Support Vector Data Description (SVDD) for analog circuits. Simulation and experiments have verified the effectiveness.