Design Of Transformer Status Monitoring System Based On Vibration Characteristics

Transformer is the most important equipment in the power system.The operation reliability of it is especially important.During the long running process,the iron core and winding of the transformer will malfunction inevitably and then lead to transformer fault.Vibration analysis,the fundamentals of it is using vibration acceleration sensor to obtain vibration signal from the surface of the transformer box,in turn judging the state of winding and iron core according to the characteristics of measured signal.This method is characterized by simple installation,economical efficiency and non-electric connection with the whole power system.In this thesis,the traditional detection methods with respect to transformer winding and iron core and their advantages and disadvantages are discussed.The research status and present shortcomings of vibration method in transformer detection are focused.From the above,the emergence and transmission of transformer vibration are analyzed.An equivalent model of the winding vibration is established by analyzing the stress of the transformer windings.The relationship among the winding vibration acceleration,the load current and frequency and the effect of the compaction state of the winding on the acceleration amplitude are reached as well.The cause of the vibration in iron core and the relationship between the acceleration of the core vibration and the applied voltage arc studied.The transformer vibration monitoring system is designed,which contains two parts:the software and hardware.The hardware part includes the main chip module,USB interface module and voltage regulator circuit module.The software adopted LabVIEW to design and then applied it in single-phase transformer vibration monitoring experiment to verify the feasibility of the monitoring system.Then the monitoring system is used in three-phase transformers to analyze the vibration conditions of transformers at different phases and locations.The effectiveness of the monitoring system is verified by the vibrations of the transformer core and winding obtained from no-load and load experiments respectively.In the end,the thesis introduces a blind separation algorithm for transformer vibration signals.It verifies the feasibility of the algorithm in separating transformer vibration signal from the perspective of simulation.Then vibration signal from the transformer box surface is blindly separated by the algorithm to receive the vibration signals of the transformer winding and the iron core.Thus,the state of the transformer inner winding and the iron core can be measured.