PMSG MPPT Control Study Without Mechanical Sensor In Wind Power System

In recent years, Direct-drive pe rmanent m agnet synchronous wind power generation has become a new hotspot in wind power technology research because of its high reliability, high proficiency and non-direct grid coupling character. This paper intend to introduce a new m ethod to re alize the m aximum po wer point tracking (MPPT)without any m echanical sensor. sensorless control technology reduces cost, increases system reliability, so it h as been a focus of motor control r esearch for a period of time. The mechanical sensors mentioned here refer to two kinds of sensor , the first one is rotor position sensor . in th e permanent magnet motor vector con trol system that calls f or high performances, rotor position information is imperative for orientation. The second one is wind speed se nsor or anemoscope. The mostly applied method for MPPT directly uses wind inform ation to calculate the rotational speed of the turbine. Obviously simplicity is the best advantage of this method while it has its considerable drawbacks. Papers concer ning PMSM sensorless control m ethod and MPPT sensorless control m ethod have been widely published respectively , however, very few paper has tried to combine the two since this combination is more difficult than just one plus one.MPPT has been widely used in solar en ergy and VSCF(Variable Speed Constant Frequency) wind power genera tion. And it is a core issue in the renewable energy optimum exploitation. Based on the wind turbine m odel within Matlab, this paper firstly introduces several m ethods to re alize MPPT with and without wind speed sensor, and focuses on the so called optimum torque m ethod. Then, secondly introduces the vecto r control as well a s SVPWM. Moreover, this paper also uses Particle Swarm Optimization (PSO) method to optimize the PID param eters, and the step response of the close loop based on th ese PID parameters shows better dynamic performance than by traditional method.The key focus of this paper is about PMSG (perm anent magnet synchronous generator) position sensorless control by using sliding mode observer (SMO)and EKF observer.Traditional SMO uses current as the state variables to provide the state equations. And it uses EMF to calculate the position information. The steering function of SMO is a sign function, so it causes extreme chattering phenomenon. The result cannot be directly used unless a low-pass filter is coupl ed to it. But the filte r will cause angular delay and amplitude reduction.To solve the delaying problem , th is paper invents a dynam ic angular compensation method to compensate the delay. At the sam e time, with this m ethod, the rotational speed can be calculated through the process and without using angular derivative which greatly causes fluctuation. Th e stability of this m ethod is proved by using Lyapunov principle. The system is modeled and the result is analyzed in Matlab environment.The Kalm an observer is based on the Kalm an f ilter. Th e state equ ation is provided based on the two-phase stationary orientation using currents, angle, and rotational speed as the s tate variables. And it us es currents error as the feedback to estimate the angle and rotational speed by applying the discrete EKF. The simulation shows a very delightful result.