| The permanent magnet synchronous motor,in which permanent magnet was used instead of excitation winding and other components,greatly simplifies the motor structure.Meanwhile,it has the advantages of high power density,high power factor and high operation reliability,and has become the mainstream motor type used in various fields.The interior permanent magnet synchronous motor is widely used in weapon equipment,industrial application and other fields because of its strong overload ability.In PMSM control system,the use of position sensor leads to the increase of system cost and the decrease of system reliability,especially limits its application in special environment.Therefore,in order to improve the cost performance and reliability of the motor control system,and make it suitable for worse environment,IPMSM sensorless control technology has become a research hotspot in recent years.The model method based on back-EMF has good observation performance in the middle and high speed domain,and the sliding mode observer is the most widely used because of its strong immunity,fast dynamic response and insensitivity to the change of its own parameters.However,due to factors such as the nonlinearities of the inverter and the phase lag of the low-pass filter,the problem of rotor position observation error exists in the application of the sliding mode observer.Those error factors will greatly affect the accuracy of rotor position observation,and ultimately affect and limit the performance of sensorless control system.To solve this problem,the main work and achievements of this paper are as follows:1.The SMO model based on the extended back-EMF and the normalized orthogonal phase-locked loop are established.The estimation error of the traditional rotor position observer is analyzed and its error model is established.2.A new rotor position observer based on proportional resonant-sliding mode control surface is proposed.The stability of the observer is proved theoretically by Lyapunov stability criterion.The simulation and experiment show that the method can suppress all harmonics except the fundamental wave and effectively reduces the estimation rotor position pulsation error.3.The DC offset errors in both the aforementioned new observer and the conventional observer are analyzed,and an improved rotor position observer based on proportional resonant filter is proposed.The complete transfer function of the proposed improved observer is derived and established,and the parameters of the proportional resonant filter are designed according to its influence on the global performance.It is also verified by simulation and experiment that the proposed improved observer can effectively suppress the harmonic pulsation of rotor position estimation and significantly reduce the DC offset error.Finally,comparative experiments are carried out on the 1.5KW IPMSM sensorless driving control platform.The experimental results show that the proposed improved observer has a good performance of position estimation error suppression. |
