Mohit Sinha

Uncovering Droop Control Laws Embedded Within the Nonlinear Dynamics of Van der Pol Oscillators

This paper examines the dynamics of power-electronic inverters in islanded microgrids that are controlled to emulate the dynamics of Van der Pol oscillators. The general strategy of controlling inverters to emulate the behavior of nonlinear oscillators presents a compelling time-domain alternative to ubiquitous droop control methods which presume the existence of a quasistationary sinusoidal steady state and operate on phasor quantities. We present two main results in this paper. First, by leveraging the method of periodic averaging, we demonstrate that droop laws are intrinsically embedded within a slower time scale in the nonlinear dynamics of Van der Pol oscillators. Second, we establish the global convergence of amplitude and phase dynamics in a resistive network interconnecting inverters controlled as Van der Pol oscillators. Furthermore, under a set of nonrestrictive decoupling approximations, we derive sufficient conditions for local exponential stability of desirable equilibria of the linearized amplitude and phase dynamics.

Synthesizing Virtual Oscillators to Control Islanded Inverters

Virtual oscillator control (VOC) is a decentralized control strategy for islanded microgrids where inverters are regulated to emulate the dynamics of weakly nonlinear oscillators. Compared to droop control, which is only well defined in sinusoidal steady state, VOC is a time-domain controller that enables interconnected inverters to stabilize arbitrary initial conditions to a synchronized sinusoidal limit cycle. However, the nonlinear oscillators that are elemental to VOC cannot be designed with conventional linear-control design methods. We address this challenge by applying averaging- and perturbation-based nonlinear analysis methods to extract the sinusoidal steady-state and harmonic behavior of such oscillators. The averaged models reveal conclusive links between real- and reactive-power outputs and the terminal-voltage dynamics. Similarly, the perturbation methods aid in quantifying higher order harmonics. The resultant models are then leveraged to formulate a design procedure for VOC such that the inverter satisfies standard ac performance specifications related to voltage regulation, frequency regulation, dynamic response, and harmonic content. Experimental results for a single-phase 750 VA, 120 V laboratory prototype demonstrate the validity of the design approach. They also demonstrate that droop laws are, in fact, embedded within the equilibria of the nonlinear-oscillator dynamics. This establishes the backward compatibility of VOC in that, while acting on time-domain waveforms, it subsumes droop control in sinusoidal steady state.

Nonlinear supersets to droop control

This paper offers a suite of extensions to Virtual Oscillator Control - a time-domain control method for islanded inverters whereby they are controlled to emulate the dynamics of weakly nonlinear limit-cycle oscillators. First, we develop a coordinate transformation to derive the PWM switching signals from the oscillator dynamic states in a manner that allows the inverter terminal-voltage amplitude and frequency to be traded off against a parametric linear combination of average active and reactive power. Additionally, we compare the time-domain performance of VOC to droop control for parallel connected inverters in two cases: synchronization from a cold start and inverter addition. Finally, with a view towards developing output filter-design strategies as well as outlining strategies for grid-connected operations, we derive a sufficient condition for a Virtual-Oscillator-controlled inverter to be entrained to a stiff voltage source.

Virtual Oscillator Control subsumes droop control

In this paper we examine the amplitude and phase dynamics of power-electronic inverters in islanded microgrids that are controlled to emulate the dynamics of a class of weakly nonlinear Liénard-type oscillators. The general strategy of controlling inverters to emulate the behavior of Liénard-type oscillators is termed Virtual Oscillator Control (VOC), and it presents a compelling time-domain alternative to ubiquitous droop control methods which linearly trade off voltage frequencies and magnitudes with active and reactive power injections. In comparison to droop control, which assumes a priori that the network operates in a quasi-stationary sinusoidal steady state, VOC is a time-domain control strategy that globally stabilizes a desired sinusoidal steady state. The main, and somewhat surprising, result of this paper is that-when reduced to the sinusoidal steady state-the VOC dynamics correspond to those of droop control. Hence, VOC is a globally stabilizing control strategy that can deal with higher-order harmonics and includes droop control in the harmonic steady state. The results are intriguing, in that they suggest that droop control laws can be recovered from averaging the complex dynamics of a class of weakly nonlinear limit-cycle oscillators.

Synchronization of Liénard-type oscillators in uniform electrical networks

This paper presents a condition for global asymptotic synchronization of Liénard-type nonlinear oscillators in uniform LTI electrical networks with series R-L circuits modeling interconnections. By uniform electrical networks, we mean that the per-unit-length impedances are identical for the interconnecting lines. We derive conditions for global asymptotic synchronization for a particular feedback architecture where the derivative of the oscillator output current supplements the innate current feedback induced by simply interconnecting the oscillator to the network. Our proof leverages a coordinate transformation to a set of differential coordinates that emphasizes signal differences and the particular form of feedback permits the formulation of a quadratic Lyapunov function for this class of networks. This approach is particularly interesting since synchronization conditions are difficult to obtain by means of quadratic Lyapunov functions when only current feedback is used and for networks composed of series R-L circuits. Our synchronization condition depends on the algebraic connectivity of the underlying network, and reiterates the conventional wisdom from Lyapunov- and passivity-based arguments that strong coupling is required to ensure synchronization.

Design of optimal coupling gains for synchronization of nonlinear oscillators

This paper develops a structured optimal-control framework to design coupling gains for synchronization of weakly nonlinear oscillator circuits connected in resistive networks with arbitrary topologies. The oscillators are modeled as weakly nonlinear Liénard-type circuits, and the coupling gain amounts to the current gain which scales the output current of the oscillator. The structured optimal-control problem allows us to seek a decentralized control strategy (equivalently, a diagonal feedback matrix) that precludes communications between oscillators. To this end, a sparsity-promoting optimal control algorithm is developed to tune the optimal diagonal feedback-gain matrix with minimal performance sacrifice. This involves solving an ℋ2 optimal control problem with ℓ1 regularization by applying the alternating direction method of multipliers (ADMM). Simulation studies with application to voltage regulation in islanded networks composed of power-electronic inverters are provided to validate the approach.

Comparison of virtual oscillator and droop control

Virtual oscillator control (VOC) and droop control are distinct methods to ensure synchronization and power sharing of parallel inverters in islanded systems. VOC is a control strategy where the dynamics of a nonlinear oscillator are used to derive control states to modulate the switch terminals of an inverter. Since VOC is a time-domain controller that reacts to instantaneous measurements with no additional filters or computations, it provides a rapid response during transients and stabilizes volatile dynamics. In contrast, droop control regulates the inverter voltage in response to the measured average real and reactive power output. Given that real and reactive power are phasor quantities that are not well-defined in real time, droop controllers typically use multiplicative operations in conjunction with low-pass filters on the current and voltage measurements to calculate such quantities. Since these filters must suppress low frequency ac harmonics, they typically have low cutoff frequencies that ultimately impede droop controller bandwidth. Although VOC and droop control can be engineered to produce similar steady-state characteristics, their dynamic performance can differ markedly. This paper presents an analytical framework to characterize and compare the dynamic response of VOC and droop control. The analysis is experimentally validated with three 120 V inverters rated at 1kW, demonstrating that for the same design specifications VOC is roughly 8 times faster and presents almost no overshoot after a transient.

Decentralized interleaving of paralleled dc-dc buck converters

We present a decentralized control strategy that yields switch interleaving among parallel-connected dc-dc buck converters. The proposed method is based on the digital implementation of the dynamics of a nonlinear oscillator circuit as the controller. Each controller is fully decentralized, i.e., it only requires the locally measured output current to synthesize the pulse width modulation (PWM) carrier waveform and no communication between different controllers is needed. By virtue of the intrinsic electrical coupling between converters, the nonlinear oscillator-based controllers converge to an interleaved state with uniform phase-spacing across PWM carriers. To the knowledge of the authors, this work presents the first fully decentralized strategy for switch interleaving in paralleled dc-dc buck converters.

Proceedings of the American Control Conference

This paper considers a collection of networked nonlinear dynamical systems, and addresses the synthesis of feedback controllers that seek optimal operating points corresponding to the solution of pertinent network-wide optimization problems. Particular emphasis is placed on the solution of semidefinite programs (SDPs). The design of the feedback controller is grounded on a dual ε-subgradient approach, with the dual iterates utilized to dynamically update the dynamical-system reference signals. Global convergence is guaranteed for diminishing stepsize rules, even when the reference inputs are updated at a faster rate than the dynamical-system settling time. The application of the proposed framework to the control of power-electronic inverters in AC distribution systems is discussed. The objective is to bridge the time-scale separation between real-time inverter control and network-wide optimization. Optimization objectives assume the form of SDP relaxations of prototypical AC optimal power flow problems.