Research On The Control Strategy Of Inverter Based On Repetitive Control And PI

With people's requirement of inverter waveform quality improved, waveform control technique for inverter has been greatly development in topology, modulation strategy and control strategy. Among them, the inverter waveform control strategy is particularly significant. However, most of the current control strategies can only optimize a certain aspect of the system such as dynamic, steady-state and so on, and can't take into account all. Therefore, the system is unable to adapt to the complex working environment. The inverter waveform control strategy needs to be further improved.Firstly, the physical model of three-phase inverter is established. Combining with the actual parameters, the characteristic of the control object is derived. At the same time, according to the classification of common load, the influence of each type of load disturbance on the waveform quality is analyzed from the direction of harmonic wave. Secondly, the thesis designed control systems and respectively derived the both control performance between instantaneous PI control and direct repetitive control. The parameters are designed by Bode diagram method and the wave quality of both control schemes are compared by simulation. It is concluded that the instantaneous PI control is faster than direct repetitive control on dynamic state, and the direct repetitive control is excel to instantaneous PI control on steady state. This conclusion is proved by the experiment of small power prototype.This thesis proposes a modified repetitive controller based on N/6 and the delay time of the direct repetitive control is replaced by the 1/ 6 fundamental wave period. Through theoretical analysis, it is concluded that this controller can effectively improve the dynamic performance of repetitive control. However, the improved controller introduces the problem of unbalanced load and internal model division precision. The solutions are proposed based on the symmetrical component method and the Lagrange interpolation method. The improved repetitive controller and the solution are synthesized, and the structure of the improved compound repetitive controller is given. The simulation results show that the proposed controller has fast dynamic performance and good static performance in the complex working environment. Finally, this thesis presents a parallel type of repetitive controller, and further improvements are made on the dynamic performance of the controller, which can be proved by theoretical analysis.