Power switchgear

as an important component of power systems

is widely distributed across all levels of the grid

and its mechanical performance significantly affects the safe and reliable operation of equipment. Traditional monitoring methods

such as grating displacement measurement

although possessing high precision

have limitations in environmental adaptability and deployment flexibility in practical applications. To address the challenge of monitoring the mechanical performance of circuit breaker operating mechanisms

this paper proposes a novel event-stream-driven fault detection method based on a dynamic vision sensor (DVS). This method collects event-stream data during opening and closing operations under healthy and faulty conditions

and combines an event-based time-windowed voxel reconstruction (ETVR)

optical flow target tracking (Lucas-Kanade)

and template matching (NCC method) techniques to extract key mechanical motion parameters with high precision. The experimental results show that the timing errors for both closing and opening operations under healthy and faulty conditions are all below 2%

and the method can be adopted to effectively identify spring fatigue faults and achieve early degradation detection of the mechanism. This study provides an innovative technical approach for intelligent mechanical condition monitoring and diagnosis of switchgear

significantly improving the accuracy

real-time capability

and automation level of monitoring.