Search

This article proposes a novel optimal hybrid control framework to improve the voltage profile of highly unbalanced Distribution Grids (DGs) by coordinating the injection of reactive power from multiple off-board Electrical Vehicles (EVs) chargers. It exploits the time-synchronized measurements advantages, centralized control, and local controllers embedded into a hierarchical and scalable scheme, coordinated by the distribution system operator (DSO). The proposed framework is aimed to increase the modern electrical networks’ flexibility, reliability, and stability. It is powered by the Eigensystem Realization (ER)-based system identification technique, a Linear Quadratic Gaussian (LQG) controller, the Distribution-level Phasor Measurement Units (D-PMUs), and a new aggregator agent that handles the EV chargers power injection to precisely and timely regulate the dynamic voltage response in a sub-area or the entire DG. The robustness and feasibility of the proposal are demonstrated by employing simulated scenarios with unbalanced faults, latency in communications, and highly unbalanced loads conditions on the IEEE 13 and 123 nodes test feeders, compensating the voltage variations accurately in less than 205 ms. The promising outcome of this study suggests a novel application for an emerging measurement-based control system in the operation of modern active DGs.

This paper proposes a robust adaptive hybrid optimization approach taking advance of two meta-heuristic algorithms to solve the non-convex economic dispatch (ED) problem in large-scale power grids. The demanded load is continuously increasing, leading to the operators to cope with new challenges associated with the reduction of the generation costs. As a consequence of this unmeasured increasing, algorithms with flexible and soft computing capabilities for analysing large-scale power systems are required. Thus, the key idea behind the proposed hybrid optimization framework is based on the abilities of the adaptive simulated annealing (ASA) and genetic operators for evaluating the economic dispatch problem, since these accelerate the convergence and reduce the total number of evaluations. This proposal optimizes the solution of the non-convex ED problem, decreasing the generation costs and emissions, and demonstrating a better performance during its convergence characteristics. Numerical tests are carried out in two power grids, where a reduced version of the Mexican interconnected power system is considered with a 24-h wind profile, aiming to demonstrate the superiority of this work, and show the new advantages regarding other algorithms.