Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userGrid-aware Distributed Model Predictive Control of Heterogeneous Resources in a Distribution Network: Theory and Experimental Validation
327
0
0.0
(
0
)
Ask ChatGPT about the research
No Arabic abstract
In this paper, we propose and experimentally validate a scheduling and control framework for distributed energy resources (DERs) that achieves to track a day-ahead dispatch plan of a distribution network hosting controllable and stochastic heterogeneous resources while respecting the local grid constraints on nodal voltages and lines ampacities. The framework consists of two algorithmic layers. In the first one (day-ahead scheduling), we determine an aggregated dispatch plan. In the second layer (real-time control), a distributed model predictive control (MPC) determines the active and reactive power set-points of the DERs so that their aggregated contribution tracks the dispatch plan while obeying to DERs operational constraints as well as the grids ones. The proposed framework is experimentally validated on a real-scale microgrid that reproduces the network specifications of the CIGRE microgrid benchmark system.
rate research
Read More
In this paper, we consider a stochastic Model Predictive Control able to account for effects of additive stochastic disturbance with unbounded support, and requiring no restrictive assumption on either independence nor Gaussianity. We revisit the rather classical approach based on penalty functions, with the aim of designing a control scheme that meets some given probabilistic specifications. The main difference with previous approaches is that we do not recur to the notion of probabilistic recursive feasibility, and hence we do not consider separately the unfeasible case. In particular, two probabilistic design problems are envisioned. The first randomization problem aims to design textit{offline} the constraint set tightening, following an approach inherited from tube-based MPC. For the second probabilistic scheme, a specific probabilistic validation approach is exploited for tuning the penalty parameter, to be selected textit{offline} among a finite-family of possible values. The simple algorithm here proposed allows designing a textit{single} controller, always guaranteeing feasibility of the online optimization problem. The proposed method is shown to be more computationally tractable than previous schemes. This is due to the fact that the sample complexity for both probabilistic design problems depends on the prediction horizon in a logarithmic way, unlike scenario-based approaches which exhibit linear dependence. The efficacy of the proposed approach is demonstrated with a numerical example.
In this paper, we present a data-driven secondary controller for regulating to some desired values several variables of interest in a power system, namely, electrical frequency, voltage magnitudes at critical buses, and active power flows through critical lines. The power generation system is based on distributed energy resources (DERs) interfaced with either grid-forming (GFM) or grid-following (GFL) inverters. The secondary controller is based on online feedback optimization leveraging the learned sensitivities of the changes in the system frequency, voltage magnitudes at critical buses, and active power flows through critical lines to the changes in inverter active and reactive power setpoints. To learn the sensitivities accurately from data, the feedback optimization has a built-in mechanism for keeping the secondary control inputs persistently exciting without degrading its performance. The feedback optimization also utilizes the learned power-voltage characteristics of photovoltaic (PV) arrays to compute DC-link voltage setpoints so as to allow the PV arrays to track the power setpoints. To learn the power-voltage characteristics, we separately execute a data-driven approach that fits a concave polynomial to the collected power-voltage measurements by solving a sum-of-squares (SoS) optimization. We showcase the secondary controller using the modified IEEE-14 bus test system, in which conventional energy sources are replaced with inverter-interfaced DERs.
This paper proposes an off-line algorithm, called Recurrent Model Predictive Control (RMPC), to solve general nonlinear finite-horizon optimal control problems. Unlike traditional Model Predictive Control (MPC) algorithms, it can make full use of the current computing resources and adaptively select the longest model prediction horizon. Our algorithm employs a recurrent function to approximate the optimal policy, which maps the system states and reference values directly to the control inputs. The number of prediction steps is equal to the number of recurrent cycles of the learned policy function. With an arbitrary initial policy function, the proposed RMPC algorithm can converge to the optimal policy by directly minimizing the designed loss function. We further prove the convergence and optimality of the RMPC algorithm thorough Bellman optimality principle, and demonstrate its generality and efficiency using two numerical examples.
In Part I, a method for the Harmonic Power-Flow (HPF) study of three-phase power grids with Converter-Interfaced Distributed Energy Resources (CIDERs) is proposed. The method is based on generic and modular representations of the grid and the CIDERs, and explicitly accounts for coupling between harmonics. In Part II, the HPF method is validated. First, the applicability of the modeling framework is demonstrated on typical grid-forming and grid-following CIDERs. Then, the HPF method is implemented in Matlab and compared against time-domain simulations with Simulink. The accuracy of the models and the performance of the solution algorithm are assessed for individual resources and a modified version of the CIGRE low-voltage benchmark microgrid (i.e., with additional unbalanced components). The observed maximum errors are 6.3E-5 p.u. w.r.t. voltage magnitude, 1.3E-3 p.u. w.r.t. current magnitude, and 0.9 deg w.r.t. phase. Moreover, the scalability of the method is assessed w.r.t. the number of CIDERs and the maximum harmonic order ($leqslant$25). For the maximum problem size, the execution time of the HPF method is 6.52 sec, which is 5 times faster than the time-domain simulation. The convergence of the method is robust w.r.t. the choice of the initial point, and multiplicity of solutions has not been observed.
This brief introduction to Model Predictive Control specifically addresses stochastic Model Predictive Control, where probabilistic constraints are considered. A simple linear system subject to uncertainty serves as an example. The Matlab code for this stochastic Model Predictive Control example is available online.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
