Research On Model Predictive Control Of PWM Rectifier

With the continuous progress of China’s modernization process,the precision of various industrial equipment and household appliances is also getting higher and higher.In order to improve the production efficiency and prolong the service life of the equipment,people put forward higher requirements for power quality.However,both in the industrial and daily life field there are a lot of non-linear load,which produce a large number of harmonics,resulting in low power factor and bring adverse effect on electrical equipment.Compared with diode and other uncontrollable rectifier,as a "green" power conversion device,PWM rectifier not only can realize the sinusoidal current of the network side,but also can achieve DC side voltage stability and controllable,which could eliminate harmonic source from the root to reduce power losses and improve the operating efficiency of electrical equipment.Based on the structure of two-level three-phase voltage source PWM rectifier,in this paper,the control strategy of PWM rectifier is studied.In this paper,the mathematical model of three-phase voltage source PWM rectifier is deduced in detail,and its control strategy is studied deeply.As a relatively mature control strategy,the control strategy based on grid voltage orientation can achieve good steady-state performance,but its dynamic performance is not ideal and the design and adjustment of PI parameters are troublesome.In order to improve the dynamic performance of the rectifier and simplify the design of the controller parameters,this paper introduces the model predictive control strategy.Single-vector FCS-MPC control strategy is simple and its dynamic performance is good,but its switching frequency is not fixed,the equivalent switching frequency is low and steady-state performance is poor.The steady-state performance of the two-vector FCS-MPC is much better than that of the single-vector FCS-MPC,but its dynamic response is worse.FCS-MPC takes eight vectors into the rectifier mathematical model when calculating the optimal vector,which greatly increases the burden on the processor,and it’s not conducive to the improvement of system performance.Aiming at this problem,this paper presents a simplified algorithm,which does not need to calculate the predicted current of eight vectors,thus reducing the computing time of the processor.Whether the single-vector FCS-MPC or two-vector FCS-MPC,power transistor’s switching frequency is not constant.In order to achieve constant switching frequency,improve the steady-state performance of the rectifier,this paper proposes a constant-switching-frequency model predictive control strategy,which calculates the voltage vector acting at the next sampling time,and converts the voltage vector into the corresponding switching signal by SVPWM modulation algorithm to realize the fixed frequency control.Aiming at the shortcoming of the model predictive control algorithm which is susceptible to the mismatch of model parameters and the influence of external disturbance,this paper combines the extended state observer with the model predictive control.The model error caused by the parameter mismatch and the external disturbance are observesd through the observer,then the observed results are fed forward and the mathematical model of the system is modified in real time to eliminate the influences of parameter mismatch and the external disturbance and realize the robust control of the prediction algorithm.In this paper,the observation error of the extended state observer is analyzed.The results show that the disturbance observed by the extended state observer is very close to the actual value,and it can be seen that it is appropriate to observe the disturbance by using the extended state observer.In order to verify the effectiveness of the algorithm,this paper has done the relevant simulation and experiment,simulation and experimental results have verified the effectiveness of the proposed method.