Design And Research Of Fault Automatic Monitoring System Of Outdoor High-voltage Circuit Breaker Motor Operating Mechanism

Outdoor high-voltage circuit breaker motor operation mechanism is an important part of the power distribution system,Its state directly determines whether the power system is reliable,do a good job of related operation and maintenance is an important condition and foundation to ensure the safe and stable operation of the power grid.At present,in areas with serious partial loads,motor operating mechanisms often have the relevant equipment performance indicators are too low,and the failure cannot be identified and detected.In order to solve the problem of fault diagnosis of the current high-voltage circuit breaker motor operating mechanism,this project designs a complete online monitoring and fault detection system of the operating mechanism.The system can effectively identify nearly 40 kinds of motor and control system faults in the operating mechanism.In the motor fault diagnosis,starting from the electromagnetic theory of the motor,the faults such as broken strip,eccentricity,and short circuit between turns of the motor rotor are theoretically analyzed.Aiming at the problem that the measured signal contains other interference signals and its fault characteristic components are difficult to be identified,this thesis uses empirical mode decomposition method to reduce noise and filter the accurate signal components,and then through the spectrum analysis of the failure data of the motor,the characteristic components reflecting the faults such as short circuit between turns of the motor coil,rotor breakage,and motor eccentricity are obtained,which provides a basis for fault diagnosis.In the fault diagnosis part of the control system,the high-voltage circuit breaker motor operation mechanism is only closed several times a year,which has the characteristics of short working time and long standby time.Traditional current monitoring methods cannot monitor the condition of equipment in real time during non-working hours.In this thesis,the main control chip injects high-frequency detection signals into the original split-close control loop in real time,monitors each component node,and realizes real-time monitoring and fault diagnosis of equipment during static working hours.In view of the problem of insufficient use of existing features and low prediction efficiency in the fault diagnosis algorithm,this thesis first adopts the polynomial feature derivation method to encode the nonlinear law in the feature space,which greatly improves the utility of the features,and secondly,proposes a fault classification model of power system based on ensemble learning,which divides the identification process into three layers,and each layer is extended to a separate fault classification algorithm.Compared with the traditional algorithm,the algorithm improves accuracy,precision,recall and F1 value.Finally,STM32F407 was used as the main control chip to complete the design of the system’s software and hardware circuits.In terms of the hardware of the system,the three-phase motor current and voltage acquisition channel circuit,characteristic signal monitoring channel circuit,detection signal channel circuit,isolated power supply and other circuits are selected and designed in detail.In terms of system software,the embedded system using u C/OS-II is used to implement FFT algorithm,data storage,WIFI data communication,AD sampling,feature extraction,and fault diagnosis.A complete fault monitoring and diagnosis system has been established,which can effectively identify nearly 40 fault categories and achieve the overall goal plan.