Research On Transient Analysis Method Of Broken Rotor Bar Of Squirrel Cage Motor

As an important electromechanical energy conversion device,squirrel cage motor is widely used in the industrial and agricultural production in China.The operation of squirrel cage motor with "disease" will not only cause damage to itself,but also cause the equipment to stop running when the fault is serious,so that the production of the enterprise will be completely paralyzed,which will bring harm and inestimable economic loss to the enterprise personnel and the society.The fault operation of squirrel cage motor is also one of the important reasons to reduce its working efficiency and waste energy.Therefore,the running status of squirrel cage motor State monitoring and fault diagnosis are of practical significance in ensuring continuity of production,reducing motor drive and system maintenance costs.Rotor bar breakage is a special fault type of squirrel cage motor,and it is one of the hot topics in the field of motor fault diagnosis.The classical MCSA method is an effective method to diagnose the broken bar fault of cage motor rotor,but it can not achieve the expected diagnostic effect under some special conditions.For example,when the motor is in the state of no-load operation and the transient state of the speed of the power or load fluctuates greatly,the fault characteristic frequency will be annihilated by the leakage of the fundamental frequency spectrum and the diagnosis spectrum will be blurred.In this way,the MCSA method is powerless to diagnose the broken bar fault of rotor.Therefore,this paper aims at the applicability of the classical MCSA method mentioned above.Under the condition of the transient operation of squirrel cage motor,the fault of broken rotor bar is studied.The theoretical analysis shows that the relationship between the instantaneous frequency of harmonic and slip in the stator current park vector mode squared signal is fixed.Based on this theory,a fault diagnosis method for rotor broken bars based on instantaneous frequency analysis of fault characteristic harmonics is proposed.The main idea of this method is to extract the fault characteristic harmonics which directly reflect the fault state of rotor bar by using the spectrum separation function of Park vector transform and the low-pass filtering function of discrete wavelet transform.Based on the concept of instantaneous frequency,the time-frequency evolution law of fault characteristic harmonics is described and graphically.The fault characteristic curve is compared with the ideal characteristic to diagnose rotor broken bar fault.However,the proposed diagnosis method can only judge whether the fault occurs qualitatively or not,and can not give the result of judging the severity of the fault.As an improvement of the above methods,a diagnosis method based on WVD distribution graphics and WVD energy calculation is proposed.In the proposed method,according to the correlation between the graphical fault characteristic curve on the energy-slip plane and the theoretical characteristics,the fault severity of the rotor broken bar of the squirrel cage motor is qualitatively judged,and the quantitative analysis is carried out according to the WVD energy calculation.The experimentalresults show that when the number of rotor broken bars is not too large,the corresponding harmonic WVD energy increases with the deepening of fault severity.The results show that the proposed method can correctly judge the motor with different fault severity.Due to the ideal time-frequency aggregation of WVD,the fault characteristics of the proposed diagnosis method are clear and the proposed method has higher reliability.