Research On Fault Detection And Identification Of Broken Rotor Bars Of Induction Motor Under Dynamic Speed Regulation

The variable frequency speed regulation system composed of inverter and induction motors(IMs)have been widely used in actual industrial production due to the advantages of good reliability,high efficiency and energy saving.However,adverse industrial environment cause motor failures to be inevitable.Among them,the broken rotor bars(BRBs)fault is a common type among various motor faults,accounting for about 10%.Once a fault occurs,it is easy to cause major production safety accidents and cause incalculable economic losses.Therefore,for the safety and efficiency of production,it is of great significance to detect and identify BRBs faults under variable frequency speed regulation.Compared with the line-fed IMs,the BRBs fault detection and identification of IMs under variable frequency speed regulation are more challenging.On the one hand,nonlinear and time-varying components are contained in the BRBs-related harmonics.With the change of operation mode,the signal frequency and amplitude will change in a wide range.It is difficult for traditional detection methods to extract fault features.On the other hand,when the inverter gets involved,a large amount of harmonic noise follows.Among them,there are many harmonics with similar BRBs-related harmonics,which are easy to confuse the fault feature,resulting in false detection of BRBs.The main research of the thesis can be outlined as follows:(1)In order to overcome the problem that the BRBs fault feature is difficult to extract under dynamic speed regulation condition,a novel approach called Adaptive Window Short-time Esprit(AWSTE)is developed.Firstly,the rotating magnetic field theory and the mechanical characteristic curve of IM are used to analyze BRBs fault mechanism under the condition of variable frequency speed regulation.It is concluded that the fundamental envelope in the stator current can be used as the BRBs fault feature.Its frequency remains unchanged under constant load.On this basis,the short data sliding window is used to cut the non-stationary signal into several segments.Hence,the current signal in each window is approximately stable.Then,Esprit is used to estimate the fundamental frequency of the signal in each window.Based on the relationship between frequency error and window length,the window length is adjusted adaptively to obtain higher precision frequency values.Next,the frequency value is combined with the least squares method to calculate the fundamental amplitude of each sliding window.The fundamental amplitude sequence is formed,which is BRBs fault feature – the fundamental envelope.Finally,experimental results demonstrate the validity of the proposed approach.(2)In order to accurately identify BRBs faults,a BRBs fault identification method based on interactive identification and speed measurement is proposed.In this method,dynamic parameter identification is added on the basis of model reference adaptive system(MRAS).In every fixed cycle,the stationary segment data is used to identify parameters and update the model parameters in MRAS,which improves the accuracy of speed estimation.Based on the rotational speed,the similarity between the fundamental envelope and the theoretical fault characteristic signal(constructed based on the speed frequency)is compared with the Pearson correlation coefficient to achieve the accurate identification.Simulation and experimental results show that the proposed method can accurately identify the motor speed and BRBs fault under the condition of variable frequency speed regulation.(3)In order to better apply the proposed method to engineering practice,a Lab VIEW-Based multi-functional IMs fault detection platform is designed,which combines a variety of classical detection methods with different principles.It aims to complement the advantages and disadvantages of multiple methods to detect and identify faults from different directions.After experimental tests,the virtual fault detection system can accurately and effectively detect BRBs faults.80 figures,9 tables and 86 references are included in this thesis.