Research On Motor Stator Winding Inter-turn Short-circuit Identification Based On Moving Horizon Estimation

Based on the background of deep oil recovery in the late period of Daqing oil field,this research focuses on the on-line identification of the inter-turn short-circuit fault of the stator windings of the submersible motor,which can provide a way for the prediction of the stator short circuit fault and the protection of the motor.Taking the 1.5kw induction motor in the laboratory as an example,the fault characteristics of the stator winding inter-turn short-circuit of the induction motor are discussed in detail,and the corresponding fault parameters are introduced to establish the mathematical model of the equivalent fault of the motor.Then the Lab VIEW upper computer configuration program is developed to realize real-time monitoring of inter turn short circuit fault of induction motor stator windings.Based on the mathematical mechanism analysis of the stator winding interturn short circuit of the induction motor,the equivalent circuit topology and the state space equation model of the stator winding interturn short circuit fault of the submersible motor are established by introducing the corresponding characteristic parameters of the fault winding.When the stator winding inter-turn short-circuit occurs in each phase,the fault of the corresponding phase can be detected and located by the identification of the fault parameters,and the corresponding relationship between the short circuit number of the stator winding and the fault parameters can be quantified.A parameter identification system of induction motor based on output error method is established,and an objective function of multivariable input is designed accordingly.The TR-BFGS(T-B)off-line algorithm combining the trust-region(TR)algorithm and the Broyden method(Quasi-Newton Method)is given,and the objective function is minimized off-line based on the compound two criteria.In order to solve the problem of large difference in the magnitude of each parameter in the fault model,the sensitivity function of the same order of magnitude is obtained by the optimization of parameter optimization,that is,the sensitivity function,and the calculation is simplified.In order to realize real-time tracking of induction motor parameters,on the basis of T-B off-line algorithm,the rolling time domain prediction(MHE)online algorithm is introduced.As an optimization algorithm of output error method,the problem of parameter prediction is transformed into the minimization of the objective function in the time domain.In order to verify the accuracy and applicability of the on-line fault identification of the proposed MHE algorithm,a simulation program for the simulation of the inter-turn short-circuit fault of the motor stator winding is established by using the Matlab simulation software.The results show that the identification error of stator winding inter-turn short-circuit turns is within 0.5 turns.Combined with data acquisition card to build an induction motor fault identification experiment platform based on Lab VIEW virtual instrument,and develop a Lab VIEW configuration program that can on-line detect and locate the short-circuit fault.The accuracy of the method for identifying the inter-turn short-circuit fault of the stator winding based on the model parameter prediction is verified under different working conditions.The experimental results show that the identification error of the stator winding inter-turn short-circuit number of the proposed MHE algorithm is within 4 turns.