Robust identification of the spring energy state in circuit breaker operating mechanism is of great significance for maintaining service performance. However, establishing a mapping relationship between the sampled signal and the spring energy storage state remains a key challenge limiting its widespread application. To solve this problem, this study proposes an intelligent identification method that combines Gramian angular field(GAF) and convolutional neural network(CNN) and successfully applies it to the operating mechanism of a circuit breaker. In the proposed method, GAF is used to transform the collected time-domain signal into a two-dimensional representation, which helps track the evolution process of the dynamic characteristics of the operating mechanism. The state identification experiment of the circuit breaker operating mechanism verifies the effectiveness of the proposed intelligent diagnosis method, achieving a recognition success rate close to 100.00%. This method offers a promising approach for the robust identification of the in-service state of circuit breakers.