Robust event‐triggered output feedback learning algorithm for voltage source inverters with unknown load and parameter variations

Abstract

We consider the output feedback event-triggered control of an off-grid voltage source inverter (VSI) with unknown inductance-capacitance (L − C) filter dynamics and connected load in the presence of an input disturbance acting at the inverter. Due to uncertain dynamics and unmodeled parameters in the L − C filter connected to the VSI, we use an adaptive observer to reconstruct the system's states by measuring only the voltage at the output. The control mechanism is constructed based on an impulsive actor/critic framework that approximates the cost, the event-triggered controller, and the worst case disturbance and generates the desired AC output with the least energy dissipation.We provide rigorous stability proofs and illustrate the applicability of our results through a simulation example. KEYWORDSactor/critic structures, event-triggered control, output feedback, VSI, zero-sum game INTRODUCTIONGrowing concerns about fossil-fuel reserves, energy security, and global warming have drawn a lot of attention to renewable energy resources. 1 The increasing use of such resources 2 with intermittent generations requires "smarter controllers" for balancing power consumption and generation, often through energy exchange. In an electrical utility grid, batteries exchange energy in the form of direct current (DC) electricity. Although there are different applications that can use DC directly, various types of applications require a voltage source inverter (VSI) to convert a DC voltage to an AC voltage. The VSIs are widely used for various high performance applications, such as AC motor drives, uninterruptible power supply systems, electric vehicles, reactive power compensators and active power filters, AC power supplies, and grid connected schemes. 3,4 Recently, many researchers have focused on high performance control of VSI (see the works of Vamvoudakis and Hespanha 5,6 and the references therein). One of the difficulties in designing a desired optimal performance controller for 3502

H∞ Control for Oscillator Systems With Event-Triggering Signal Transmission of Internet of Things

“…Apply the stationary condition to (16), control and disturbance compensation policies are computed as follows:…”

“…To overcome the burden of communication resource and computation bandwidth, the event-triggering mechanism was first investigated for scheduling stabilizing control tasks [10], where a controller only receives feedback states, updates its parameters and sends control signals to plant only when an event-triggering condition is violated. Inspired by the idea, several works related to event-triggered (ET) control for multiagent have been developed [11], [12], [13], [14], [16]. In [11], the ET decentralized control scheme for interconnected nonlinear systems is proposed.…”

“…For example, distributed control algorithms were devoted to consensus problems of the systems with simple models in a kinematic form or a double integrator (see [17] for more details about the applications). An ET control learning algorithm was designed for an application of voltage source inverters in [16], where the disturbance rejection policy and the optimal control policy are approximated to drive the AC output to the reference while minimizing energy loss. Unfortunately, it only applies to single systems that are not connected to the IoT network.…”

“…1) Unlike the available algorithms of distributed optimal control for nonlinear systems that are not connected to the IoT network [11], [12], [13], [14], [16], we design an algorithm for Van der Pol oscillators in the IoT network dealing with neither controlled stable limit cycles nor stable controllable equilibrium points at the origin. Especially, the system model is in the presence of unknown internal dynamics, input constraint and external disturbance.…”

Event-triggered learning

“…Recent articles proposed to improve and augment typical event-triggered state estimation (Battistelli et al 2018, Huang et al 2017, Shi et al 2014) and control algorithms (Baumann et al 2018, Narayanan & Jagannathan 2017, Vamvoudakis & Ferraz 2018, Vamvoudakis et al 2019 with data-based techniques. In these works, learning is used to approximate intractable conditional probability densities that arise in distributed problems or to obtain tractable solutions to Hamilton-Jacobi-Bellman equations that yield optimal control policies, e.g., with model-free methods such as Q-learning, or based on neural networks based.…”