| For power equipment,partial discharge is like cancer cells inside the human body.The main cause is that the metal conductors appear sharp due to the existence of air bubbles in the internal insulating material or the inability to avoid microscopic defects in the design of the structure during the production process of the power equipment.corners or burrs.Partial discharge will cause the internal insulating material to corrode due to physical and chemical effects and cause the insulating material to age and deteriorate more rapidly,eventually causing insulation breakdown and irreparable damage to the switchgear.At present,there are many methods for the detection of partial discharge.By comparing the advantages and disadvantages of several detection methods in 12 k V Ring Network Cabinet,this paper chooses uhf detection method as the research direction of PD on-line monitoring device.In order to realize the online monitoring of partial discharge of switchgear by UHF method,this paper firstly analyzes the characteristics of high frequency,wide frequency band distribution and slightly weak signal of electromagnetic wave generated by partial discharge.First,determine the signal receiving frequency band of the whole device and design the response.The band-pass filter performs hardware filtering processing on the signal,and filters out the interference signal outside the cut-off frequency.Secondly,a low-noise radio frequency amplifier is used to amplify the filtered UHF signal in two stages to improve the signal strength and the sensitivity of the device.Then,the amplified high-frequency signal is subjected to detection processing,and the envelope of the high-frequency signal is extracted through the detection diode,and finally the partial discharge pulse signal is obtained.The electromagnetic wave signal is converted into a voltage signal after being processed by the hardware circuit,so that the weak high-frequency signal becomes higher in amplitude and lower in frequency.Finally,the AD sampling and communication circuit is designed,the partial discharge signal is sampled by the independent AD sampling chip,and the sampled data is read,filtered and uploaded by the FPGA（Field-Programmable Gate Array）,so as to realize the real-time monitoring and data uploading of the insulation state of the switchgear.After sampling and analyzing the electromagnetic wave signal of the actual working environment on site,many interference signals will still be introduced in the working frequency band of the device,such as white noise interference coupled in the line,square wave signal interference generated when the wireless communication module is working,relay or Random pulse interference generated by contactor action and common narrow-band periodic interference signals,etc.Therefore,filtering the signals collected by the device is the key to improving the accuracy of partial discharge diagnosis.In this paper,the multi-layer filtering method is adopted,and four kinds of filtering algorithms with strong adaptability（adaptive threshold filtering,clustering filtering,multi-sensor filtering）are designed for the four types of interference signals.The real PD signal is preserved.Through the partial discharge simulation experiment verification,the partial discharge online detection device designed in this paper can effectively receive and sample the electromagnetic wave signal generated by partial discharge,and can filter the surrounding noise signal through the combination of software and hardware to obtain a real and accurate discharge.The pulse signal improves the accuracy of the partial discharge diagnosis of the switch cabinet. |
PDF Full Text Request