Study on fault diagnosis of CT-26 high voltage circuit breaker based on SK attention optimization Efficient Net architecture

As the key equipment in the power system, the stable operation of the circuit breaker is very important for the safety of the whole system. As the core component of the control circuit breaker switch, it is important to make accurate fault diagnosis. To realize accurate diagnosis of breaker fault, a new method of Efficient Net breaker fault diagnosis based on coordinate attention is proposed. This method firstly collects the vibration signal when the spring releases the energy, and transforms it into a two-dimensional time-frequency diagram to realize the time-frequency relationship representation of the fault of the operating mechanism under different working conditions. Subsequently, an improved SK-MBConv module was used to help and capture the location information in the image more effectively, thus further extracting rich fault features. The experimental results show that the proposed method achieves 97.7% in classification accuracy compared with other classical classification methods. This paper not only provides an efficient and accurate new method for circuit breaker fault diagnosis, but also has important practical application value for improving the safety and stability of power system.