Research On The Identification Of PMSM’s Speed And Position Without Sensors Based On The Tanh Funtion

Permanent magnet synchronous motor (PMSM) develops rapidly in recent yearsfor its simple structure and high energy density. Electrical drive system based onpermanent magnet synchronous motor has also become a hot field. Vector controlbecomes an important method in high-performance control system; however, vectorcontrol system requires real-time speed and position information as feedback. Thedetection of the speed and position depends mainly on the high-performancemechanical sensors now, but the introduction of these sensors increases the cost of thesystem, at the same time, reduces the reliability of the system, so a kind of effectivesensorless speed and position identification method becomes an important topic.The sliding mode control structure is simple, besides, it has strong robustness forthe system disturbance. The sliding mode observer based on sliding mode variablestructure also becomes an important solution for permanent magnet motor sensorlessspeed and position estimation. This paper analyzes the classic principle of slidingmode observer and introduces its application in the identification of speed andposition of the permanent magnet synchronous motor. On this basis, this paperproposes an improved sliding mode using tanh function instead of sign function forthe problem of phase delay in the classic sliding mode observer position estimation,and the stability of the proposed sliding mode observer is derived and proved.Subsequently, a combined algorithm of using the improved sliding mode and classicsliding mode is proposed to optimize the speed identification effect of the improvedsliding mode. In order to verify the effectiveness of the improved algorithm, thispaper builds a vector control system in MATLAB/Simulink simulation environment.A number of experiments under different working conditions are simulated for theimproved method. The paper complete software development through DSPsystem-level integrated development environment based on MATLAB and LabVIEWFinally, the experiment is done on the platform based on TI TMS320F2812.The system simulation and actual experimental results prove the effectiveness ofthe proposed method on the issues of phase estimation delay, what is more, the designof the observer can track the actual speed and position accurately in steady-state anddynamic process, and achieve the effect of improved method.