Design and control aspect of segmented proportional integral‐repetitive controller parameter optimization of the three‐phase boost power factor correction rectifier

Abstract

Summary This paper presented a modeling and control aspect with an offline‐based parameter optimization algorithm for effectively tuning the controller parameters of the three‐phase boost power factor correction (PFC) rectifier system. While analyzing the shortcomings of the traditional PI control with complex control transfer functions from small‐signal models and investigating the limitations of PI control on harmonic suppression from the internal model principle, a simplified form of control transfer functions based on the small‐signal model has been proposed with segmented control action. Consequently, the methodology has the advantages of both simplified control strategies and segmented PI‐repetitive control so as to achieve the performance of high‐power factor and to reduce total harmonic distortion in a broad range of load conditions. However, the segmented PI‐repetitive control contains many control parameters. Thus, the solution and optimization for tuning the controller parameters based on heuristic offline optimization technique have been used to speed up the design of control system parameters. The validity and feasibility of the proposed design and optimization are verified using a 5‐kW three‐phase PFC rectifier with digital signal processor (DSP) TMS320F28335 prototype hardware. The results were then compared with those of traditional controller designs, which show the improvements and effectiveness of rectifier performance after adopting the proposed design and optimization technique.

Performance Improvement of Three-Phase Boost Power Factor Correction Rectifier Through Combined Parameters Optimization of Proportional-Integral and Repetitive Controller

“…PI controller with dq current reference is model initially. Then the parallel structure PI-repetitive controller is used to improve THD [19], [23]. Current loop Ldid/dt and Ldiq/dt in (2) can be expressed as…”

“…In this study, to find the optimal PIR controller gains and make an effective comparison with other objective functions, min and max ranges of the controller gains are defined in Table 2. Indeed, the parameters [Kr, m, Q(z)] ranges for the repetitive controller have been set according to the investigator's analysis and recommendations [19]- [23]. Such as, Kr is a repetitive controller gain and its larger value ensures fast system response, but if it is too large, it will make the system unstable.…”

“…To verify the outer voltage loop PI parameters [KPv, KIv], (19) has been employed to test the double closed-loop stability with the same stability margins mentioned in step 3. Furthermore, the necessary and sufficient condition for the overall three-phase PFC rectifier stability is that the bandwidth (BW) of the inner current loop must be higher than the outer voltage loop [23]. Using the MATLAB step response simulation, calculate four dynamic performance criteria in the time domain, namely Mp, Ess, tr, and ts.…”

“…It is worth noting that the PI controller is not sufficient for the inner current loop to suppress the THD, particularly 5 th and 7 th order harmonics. Therefore, PI controller needs to be combined with some other control techniques, like repetitive control (RC) [19]- [23]. However, it increases the design complexity due to multi-control loops, which can hardly be handled with conventional design methods [24].…”

Archives of Electrical Engineering

“…In order to reduce the influence of the beat frequency voltage on the control system and suppress the low-order current harmonics, the method of adding a double-traction network frequency 100 Hz notch filter is adopted in the voltage signal feedback link at the DC side in references [8] and [9]. References [10], [11] and [12] use a repetitive controller in the current inner loop, which reduces the current harmonics on the grid side, but also lowers the dynamic performance of the system. Model predictive control [13] can solve the problem of slow dynamic response, is suitable for nonlinear constrained control, and has the advantage of a simple control objective [14][15][16], so it is often used in pulse-width modulation (PWM) rectifiers [17,18].…”