Vacuum circuit breakers are crucial in power systems but face challenges in fault diagnosis and health assessment due to the complexity of fault states and the limitations of traditional diagnostic methods. This study collected travel signals from vacuum circuit breakers, which revealed critical information for distinguishing fault states. Utilizing SVM, an artificial intelligence technique, the study classifies these fault states with enhanced accuracy. Additionally, the mean deviation method was employed for predicting the health status of the circuit breakers by comparing the feature points of the travel signals to those of normal conditions, thus assessing health and predicting potential faults. Both the SVM-based diagnostic and the mean deviation prediction methods demonstrated high accuracy and reliability, proving to be valuable for practical applications. This approach not only improves fault detection accuracy and timeliness but also reduces maintenance costs and increases the overall efficiency of power systems.