X. Cui, C. Keller and A. Avestruz, “Cycle-by-Cycle Digital Control of a Multi-Megahertz Variable-Frequency Boost Converter for Automatic Power Control of LiDAR,” 2019 IEEE Energy Conversion Congress and Exposition (ECCE), Baltimore, MD, USA, 2019, pp. 702-711, doi: 10.1109/ECCE.2019.8913074.
Abstract: Dynamic voltage scaling is needed to support efficient operation despite rapidly fluctuating power demand, which is exemplified in the automatic power control of LiDAR (Light Detection and Ranging) for autonomous ground and airborne vehicles. This is especially challenging in boost converters because of more complicated dynamics. A current-mode boost converter using constant-off-time (variable-frequency) is particularly advantageous because inductor current settles in one switching cycle. Digital control is needed for programmable flexibility over a wide operating range. However, variable frequency power conversion and high speed digital control are difficult to combine in a traditional digital control framework. In this paper, we apply and demonstrate in hardware a recently published theoretical framework for cycle-by-cycle digital control of variable frequency power converters. The prototype variable-frequency boost converter with a 3 MHz peak switching frequency shows an exceptionally fast dynamic response over a wide operating range.
L. Yang, X. Cui, A. Avestruz and N. Ozay, “Correct-by-construction control synthesis for buck converters with event-triggered state measurement,” 2019 American Control Conference (ACC), Philadelphia, PA, USA, 2019, pp. 1056-1063, doi: 10.23919/ACC.2019.8814937.
Abstract: In this paper, we illustrate a new correct-by-construction switching controller for a power converter with event-triggered measurements. The event-triggered measurement scheme is beneficial for high frequency power converters because it requires relatively low-speed sampling hardware and is immune to unmodeled switching transients. While providing guarantees on the closed-loop system behavior is crucial in this application, off-the-shelf abstraction-based techniques cannot be directly employed to synthesize a controller in this setting because controller cannot always get instantaneous access to the current state. As a result, the switching action has to be based on slightly out-of-date measurements. To tackle this challenge, we introduce the out-of-date measurement as an extra state variable and project out the inaccessible real state to construct a belief space abstraction. The properties preserved by this belief space abstraction are analyzed. Finally, an abstraction-based synthesis method is applied to this abstraction. We demonstrate the controller on a constant on-time buck voltage regulator plant with an event-triggered sampler. The simulation verifies the effectiveness of our controller.
X. Cui and A. Avestruz, “Switching-Synchronized Sampled-State Space Modeling and Digital Controller for a Constant Off-Time, Current-Mode Boost Converter,” 2019 American Control Conference (ACC), Philadelphia, PA, USA, 2019, pp. 626-633, doi: 10.23919/ACC.2019.8815075.
Abstract: The effective modeling and high-speed digital control of variable frequency power converters have been a longstanding challenge. In this paper, we extend and analyze with proofs for stability and performance along with calculations for robustness, a switching-synchronized sampled-state space framework for a current-mode boost converter with constant off-time. We demonstrate a controller in this framework in both simulation and real-world hardware.
A. Ramyar, X. Cui and A. Avestruz, “Two-Port Up/Down DC-DC Converter for Two-Dimensional Maximum Power Point Tracking of Differential Diffusion Charge Redistribution Solar Panel,” 2019 20th Workshop on Control and Modeling for Power Electronics (COMPEL), Toronto, ON, Canada, 2019, pp. 1-8, doi: 10.1109/COMPEL.2019.8769644.
Abstract: Maximum power point tracking (MPPT) is one of the main challenges in solar photovoltaic systems, especially under partial shading conditions in which solar panels do not receive uniform solar irradiance. The diffusion charge redistribution (DCR) technique is an approach which uses the intrinsic diffusion capacitance of solar cells together with semiconductor switches to balance the voltage among the cells. Using this technique enables us to apply MPPT with cell-level granularity using a single module-level converter. The differential DCR (dDCR) structure is a modification to the conventional DCR structure to enable differential power processing, which leads to low insertion loss because only the mismatch power is processed by every switch. The dDCR architecture has two outputs and needs a two-port converter. This paper examines a two-port up/down dc-dc converter which is capable of doing two-dimensional MPPT of dDCR solar panels. An emulator is also presented to replicate the averaged behavior of dDCR solar panels for evaluating the converter. Prototypes of the converter and the emulator were built and a coordinate ascent method was used as the two-dimensional MPPT algorithm; the capability of the converter to do two-dimensional MPPT of dDCR solar panels was demonstrated.
X. Cui, C. Keller and A. Avestruz, “A 5 MHz High-Speed Saturating Inductor DC-DC Converter Using Cycle-by-Cycle Digital Control,” 2019 20th Workshop on Control and Modeling for Power Electronics (COMPEL), Toronto, ON, Canada, 2019, pp. 1-8, doi: 10.1109/COMPEL.2019.8769620.
Abstract: The fastest transient response of dc-dc converters, which is a dominant performance criterion for dynamic voltage scaling (DVS), is fundamentally limited by the inductor current slew rate. We employ a saturating inductor with a systematic control design to increase the slew rate without introducing extra hardware. The resulting 5 MHz saturating inductor current-mode buck converter using constant-on-time (CM-COT buck converter) controlled by a high-speed cycle-by-cycle digital controller that is designed under the switching-synchronized sampled-state space (5S) control framework exhibits a 2.6 μs large-signal voltage risetime (10%-90%) with no overshoot. Furthermore, we extend the 5S control framework to this nonlinear converter plant by utilizing singular perturbation theory. The resulting controller exhibits a large-signal stability and an ability to handle a peak slew current that is 11 times the steady-state average current in a 30 W saturating inductor CM-COT buck converter prototype.
X. Cui and A. Avestruz, “A New Framework for Cycle-by-Cycle Digital Control of Megahertz-Range Variable Frequency Buck Converters,” 2018 IEEE 19th Workshop on Control and Modeling for Power Electronics (COMPEL), Padua, 2018, pp. 1-8, doi: 10.1109/COMPEL.2018.8460055.
Abstract: This paper investigates a new method to model and control variable-frequency power converters in a switching-synchronized sampled-state space for cycle-by-cycle digital control. There are a number of significant benefits in comparison to other methods including fast dynamic performance together with ease of design and implementation. Theoretical results are presented and verified through hardware and simulations of a current-mode buck converter with constant on-time. Dynamic voltage scaling is among the applications that can benefit.
R. Y. Zhang, A. Avestruz, J. K. White and S. B. Leeb, “Design of resonance damping via control synthesis,” 2015 IEEE 16th Workshop on Control and Modeling for Power Electronics (COMPEL), Vancouver, BC, 2015, pp. 1-9, doi: 10.1109/COMPEL.2015.7236464.
Abstract: Dampers are widely used in power electronics to damp resonances, in order to reduce device stress, power loss, and electromagnetic interference. In this paper we formulate the damping problem so as to expose the fundamental tradeoff between damping amplitude peaks and minimizing power dissipation, and then use a constrained optimization approach to compute optimal Pareto frontier as a function of damper order. We use the procedure to demonstrate the diminishing returns of increasing damper order using a simple filter example, and then we demonstrate the power of the method for multiport converters. In particular, we show that using the constrained optimization procedure dramatically outperforms the standard port-by-port method.