Research On Speed Sensorless Strategy Of Induction Motor Based On Observer

Compared with the DC motors,induction motors have the advantages of simple structure,robustness,durability,low price,and large power capacity under the same volume.They are very popular in industrial and agricultural production,power transmission,military fields and other industries.Status,induction motor speed control technology has also received more and more attention.With the improvement of the speed of Cpu processing,and the continuous upgrading of power electronic devices,and the introduction of a variety of advanced control strategies,the development of AC speed regulation has been promoted.Vector control technology is favored by many scholars in recent years because it can be comparable to DC motor control effects.Speed sensorless control technology with its excellent characteristics has been more and more attention by experts at home and abroad.Based on the observer technology,this paper studies the application of reduced-order and full-order observers in a speed sensorless control system.Firstly,the two main disadvantages of the reference model,namely the voltage model in the traditional model reference adaptive method are analyzed using the methods of mathematics and simulation:the parameter changing and the pure integral problem.Using the current model to be insensitive to stator resistance at low speeds,a combined reduced-order observer model based on voltage and current model was designed to seek insensitivity to parameters in different speed ranges,by the first-order low-pass filter cut-off frequency method,the pure integral problem of reduced order model is solved.Simulation experiments verify that the above design solves the limitations of the traditional method,the speed estimation effect is accurate,and the error is minimal.Then,this paper proposes a speed sensorless control strategy based on a full-order observer.Based on Lyapunov stability theory,the motor stator current error and the observed rotor flux derived the speed adaptation rate,and the pole multiple feedback matrix was configured to ensure stable operation of the speed estimation system over a wide range.However,two conditions are assumed in the speed estimation:there is no difference between the actual flux and the observer flux and negative semi-definite matrix.When the motor runs under certain conditions,some of the assumptions may not be satisfied and the system may have instability.In response to this deficiency,using the zero-pole stability characteristic of the speed estimation transfer function,a trend?_r-T_e graph of the speed estimation is plotted on the plane,and it is analyzed that when the motor runs at low speed braking,the speed estimation instability phenomenon will occur.Using the Rolls crite,a new simple and effective feedback matrix is designed to ensure the global stability of the speed estimation.When the parameters of the full-order observer are changed,there may be some deviation between the actual flux and the observation flux.With this deviation,the flux error coefficient M is derived and the influence of M on the rotor flux is analyzed.The speed adaptation rate is re-established,which reduces the sensitivity of the estimated speed to the parameter change.Simulation experiments verify the practicability of this method.