Study On Fault Diagnosis Of Circuit Breaker Based On Improved Dynamic Time Warping Algorithm

Circuit breaker is one of the most important switching devices in power system.Its reliability affects the safety and stability of power system. The switch action of the circuit breaker is driven by mechanical mechanism. Mechanical fault will lead to abnormal closing and opening of circuit breaker. Therefore, the research of fault diagnosis for mechanical mechanism of circuit breaker is of great significance to improve the maintenance level of circuit breaker and guarantee the secure and steady operation of power system.First, in this thesis, the advantages and disadvantages of traditional fault diagnosis methods for circuit breaker are analyzed. And the research with the fault diagnosis of circuit breaker opening/closing coil as object is established. Then, the characteristics of coil current is introduced, the defects of method that extracts characteristics of coil current to diagnose fault is analyzed. To solve the shortcomings of the traditional fault diagnosis methods, the improved dynamic time warping algorithm is used to diagnose the fault of circuit breaker opening/closing coil in this thesis. The principle and improved method of dynamic time warping algorithm are introduced in detail, and the steps of this algorithm for fault diagnosis are provided. In addition, in order to obtain the actual opening/closing coil current data, a detection device of circuit breaker is designed in this thesis. The hardware design and software design are accomplished. Through the actual test, the device can not only accurately acquire opening/closing coil current signal but also accurately detect mechanical characteristics of circuit breaker. Finally, in MATLAB software,the acquired current data of opening/closing coil under normal and fault state is analyzed by using improved dynamic time warping algorithm. The coil current data includes circuit breaker working on both DC operational power supply and AC operational power supply. The results show this method can accurately and detailed reflect fault information of opening/closing coil. Compared with traditional characteristic parameter method, improved dynamic time warping algorithm has better diagnosis effect and adaptability.