Model Predictive Power Control Of Single-phase PWM Rectifier Based On Phase Detection Of Reference Frequency

Because the single-phase PWM rectifier not only has a stable DC side voltage and energy can flow in both directions,but also allows the grid-side unit power factor and grid harmonic pollution to be small,which can reduce harmonics and reactive power injected into the grid and reduce pollution.With the increases of users’ demand for power electronics device,as well as the improvement of power quality requirements,the control performance of single-phase PWM rectifiers has also become stricter.Therefore,research on high-performance power control algorithms for single-phase PWM rectifiers has important theoretical significance and engineering application value.Firstly,the topology of a single-phase two-level PWM rectifier and the working principle in different switching modes are analyzed in this paper,and its mathematical model in the twophase rotating d-q coordinate system is established.Combining with the single-phase instantaneous power theory,the mathematical model of power control is derived.Traditional direct power control and finite set model prediction power control algorithm are introduced.Secondly,the reason of phase offset based on the reference frequency phase detection algorithm is analyzed in detail.Based on this,a phase detection algorithm based on the phase offset component is proposed.It can extract the sine and cosine of the phase shift angle to compensate the phase offset angle.Even if the reference frequency is not equal to the actual voltage frequency,the grid voltage phase can be accurately detected.Meanwhile,in order to improve the dynamic response of the phase detection when the frequency is abrupt,a phase detection algorithm with double compensation is further proposed.This algorithm performs advanced compensation for the low-pass filter to eliminate the phase shift of low-frequency signals through the first-order low-pass filter.And the q-axis coordinate system voltage feed is used by adjusting the compensation of the low-pass filter in real time to improved the dynamic response for phase detection.Moreover,in order to improve the accuracy of model predictive power control,an improved second-order generalized integrator is proposesd in this paper.It can effectively eliminate the various harmonics of the α-β axis coordinate system and improve the observation accuracy of instantaneous power.And aiming at the inductance sensitivity problem in the model predictive power control,the influence of the inductance parameter mismatch rate on the reactive power is quantitatively analyzed.Based on this,a reactive power compensation algorithm is proposed to eliminate the reactive power offset caused by the inductance mismatch.The simulation experiments of this algorithm and the traditional model predictive power control algorithm are carried out.The results show that the proposed algorithm can achieve the unit power factor of the grid side and lower the harmonic content,even if the inductance mismatch rate is-90% ～ 90%.The proposed algorithm reduces the dependence of model prediction on inductance parameters.Finally,the reference frequency-based phase detection algorithm and the model predictive power control algorithm under inductor mismatch is completed on hardware-in-theloop experimental platform based on RT-LAB and DSP + FPGA.