Research On Speed Sensorless Control Method Of Asynchronous Motor Based On Full-order State Observer

For vector control system of asynchronous motor, detection of the rotate speed is a necessary work in order to realize the orientation of magnetic field and the closed-loop control. According to the mathematical model of asynchronous motor, the speed signal can be calculated by mathematic way, and its precision directly influence on the degree of accuracy of magnetic field orientation. For this reason, Applying speed sensorless technology to vector control system and accurately calculating the speed signal are becoming important research hotspots in alternating current speed control field for the past few years.Existing different speed sensorless technologies have their own relative merits and the scopes of application are restricted. Aiming at this condition, using velocity sensor vector control system as a reference standard, and in comparison with model reference adaptive control(MRAC) method which is researched widely, the thesis focused researches the speed identifying method based on full-order state observer. The specific contents are as follows:(1) The principles and evolutions of asynchronous motor vector control and speed sensorless technology are summarized, then concludes and compares thoughts and applied ranges of various kinds of motor speed identifying methods to provide a theoretical basis.(2) The thesis focused researches on the principle of speed identifying method based on full-order state observer, infers its mathematical model, and takes advantage of MATLAB/SIMULINK to build the asynchronous motor vector control system simulation model based on full-order state observer. The model is simulated and its results are compared with the ones of velocity sensor vector control and typical MRAC system, in order to verify its speed identifying performance and relative merits.(3) Directing at the drawbacks of the method based on full-order state observer, and analyzing its mathematical model, the research focuses on identifying methods of the feedback gain coefficient, accordingly develops the basic full-order state observer. Based on the state error equation, the thesis makes use of Laplace transform and the Routh criterion to do a series of mathematical derivation, and obtains a expression which can confirm value range of the coefficient, thus reduces blindness of selecting the value artificially.(4) In order to avoid complicated arithmetical operation, the thesis applies particle swarm optimization(PSO) intelligent algorithm to realize self-optimization of the feedback gain coefficient, and on this basis, develops the basic PSO algorithm to obtain a better speed identifying result and improve rate of convergence of the algorithm.