| Compared with rotary induction motor(RIM),the urban transit drive system based on linear induction motor(LIM)without mechanical transmission mechanism can realize direct-drive operation,and has outstanding advantages in climbing ability,turning radius,noise,etc.However,due to the particularity of the structure,the equivalent parameters exist nonlinear severe changes.Inaccurate parameters in traction system will adversely affect operating performance.It is difficult for compensation methods based on electromagnetic modeling or offline identification to fully consider all factors accurately.Hence,it is necessary to conduct in-depth research on online parameter identification strategies.So far,the existing online parameter identification methods for LIM and the related researches on improvement of traction performance face many deficiencies,such as the lack of comprehensive consideration for the specificity of LIM and the lack of a universal solution for high-precision parameter estimation under low-switching-frequency state.In response to the above problems,this thesis takes online parameter identification as the starting point,and carries out in-depth research combined with specificity of LIM and demands of urban transit on low switching frequency and sensorless operation.The main work and innovations are as follows:(1)Firstly,this thesis expounds the changing law of main parameters in LIM,analyzes the influence of inaccurate parameters on the control system,and selects the mathematical model suitable for parameter identification research.On this basis,two low-complexity and high-reliability online parameter identification schemes are proposed for LIM.The first method only identifies the magnetizing inductance.By combining the sliding mode observer,the differential operation is eliminated and the disturbance immunity of the system is enhanced;the second method considers the changes of the magnetizing inductance and secondary resistance comprehensively and improves the accuracy of model,besides,the decoupling of the dual-parameter identification is realized to ensure the stability of the system and reduce the difficulty of tuning.(2)To enhance the discrete domain stability of the previous parameter identification method at low switching frequencies and suppress system harmonics,a high-precision parameter identification and optimal control scheme based on model predictive control with switching terms is proposed in this paper.In the sensitivity analysis,the parameter characteristic of LIM is considered.Compared with the traditional method,it can avoid the complicated improvement of the identification system and at the same time ensure the stability and reliability of the identification results.By adopting a multi-step size and parameter correction strategy,the predictive accuracy and the current harmonic performance are effectively improved.(3)To eliminate the dependence of the previous parameter identification method and traction system on speed sensor and improve the accuracy and reliability of speed control,this thesis proposes a parallel identification strategy of magnetizing inductance and speed based on full-order observer.Compared with the traditional method,on the basis of considering the rapid change of the parameters of LIM,the design principle of the feedback gain matrix is extended to ensure the stability under low-speed regeneration braking state in continuous domain and verified by the small-signal method.Meanwhile,combined with previous low-switching-frequency model predictive control,the stability under high speed in discrete domain is further enhanced.Summarizing the above research results,this thesis has formed a set of LIM traction system suitable for low-switching-frequency operation of urban transit based on the parameter identification.On the basis of guaranteeing the safe operation of the system,the complexity reduction and harmonic suppression are realized,which can effectively improve the key indicators such as efficiency and thrust. |
