Research On Order Self-separation Method And Its Application In Fault Diagnosis Of In-wheel Motor

Rotating machinery which has been widely used in various fields such as automotive,aviation,electricity,petrochemical,metallurgy and manufacturing are key equipment in modern industry,agriculture and technology.In order to ensure the safety performance,stable operation and perfect operation efficient of rotating machinery and avoid the occurrence of severe faults or accidents,the condition monitoring and fault diagnosis of rotating machinery are necessary.In field of automobiles,the common rotating machinery are engine,transmission box,electric motor(new driving source of electric vehicle: in-wheel motor)and so on.The control parameters of these equipment are volatile for they are operated in condition of variable load and rotate speed for long time,also,there exit many external interference factors,which increase the difficulty of condition monitoring and fault diagnosis.To improve the safety performance of electric vehicle driving and promote the market application of in-wheel motor driving system,the in-wheel motor is taken as research object in this paper and the rotate speed of in-wheel motor is studied in detail.The rotate speed signal of in-wheel motor is investigated based on the reference to the way that how non-stationary signal in time domain is dealt with by traditional order analysis.And the abnormal fluctuating ingredient which is caused by faults in rotate speed signal is abstracted.Also,the type and degree of faults is identified.Firstly,the principle of traditional order analysis is introduced,and a new order analysis method-order self-separation method is proposed to solve the limits of tradition order analysis.The separating process based on the order self-separation method which is able to separate the abnormal fluctuating ingredient from rotate speed signal with fault is investigated.Secondly,the algorithm which is proposed in this paper is verified based on the simulation signal with fault of single-sine,multi-sine.The in-wheel motor is regarded as a special representative of rotating machinery,and the Matlab\Simulink model of in-wheel motor in states of normal and leakage fault are established.Then,the simulation rotate speed signals of these two states are acquired.In the meantime,by order self-separation analysis of simulation rotate speed signals,the abnormal fluctuating ingredient from rotate speed signal with fault is abstracted.Further,the test bench for fault diagnosis of in-wheel motor is designed.Leakage fault of in-wheel motor is made artificially,and the real rotate speed signals of in-wheel motor in normal state and different fault degree states are collected.The validity of order self-separation method which is able to separate fault rotate speed of in-wheel motor is proved by test research.Finally,the characteristics of orders which represent fault rotate speed signal is studied based on the order information which is acquired in states of different leakage fault degree.The relationship between the order characteristics and leakage fault degree are discussed.Combining the simulation research and experimental research,the method--order self-separation method proposed in this paper is able to separate the abnormal fluctuating ingredient which is contained in the speed signal is indicated,and the feature of faults can be acquired by order spectrums of fault rotate speed signal.The identification and judgment of fault type is realized.The order self-separation method can extend the application scope of traditional order analysis.