Βέλτιστη Λειτουργία Γραμμών Διανομής Ηλεκτρικής Ενέργειας Με Υψηλή Διείσδυση Διεσπαρμένης Παραγωγής

Abstract

The development of distributed generation (based on renewable energy), energy storage technologies implemented on low and medium voltage grid, as well as telemetry systems and smart meters, resulted in the transition of modern power distribution systems from passive to "active" or "smart" grids. The need for a central distribution management scheme which incorporates control and communication systems, enabling the network operator to have an overall supervision of the grid and to coordinate effectively the devices on it, is of utmost importance.The main purpose of this paper is the development of a robust optimization software, capable of modeling any distribution power network and optimizing its operation for the next 24 hours. Initially, all different control methods implemented on distribution networks are described, varying from the local control of any device, to the centralized and coordinated control of the grid. Moreover, the advantages of this control scheme are presented in detail. The thesis refers to the positive and negative effects of distributed generation, storage devices and voltage regulators to the grid, as well as to the mathematical challenges that arise from the optimization of electrical power systems.Next, the mathematical model developed in order to simulate the distribution networks is analyzed, as well as the details concerning the design of the multi-objective optimization problem. Specifically, there are eight different objective functions, for which user selects the weight factors he desires. The optimization problem includes linear and quadratic constraints, referring to the thermal limit of the lines, the voltage limits for every node, the characteristics of each DG unit inverter, as well as constraints which describe the operation of the remaining devices.Moreover, the proposed algorithm is tested on a standard distribution network, and specifically, it is compared to the conventional voltage regulation method. Six different indices (KPI’s) were used to quantify the improvement in smart-grid scenario, for daily and annual operation, while a variety of other networks are used as case studies in order to illustrate the remaining beneficial characteristics of the algorithm. Finally, simulations of all models, as well as the necessary calculations, are all implemented in Matlab, whereas the software used for the optimization is IBM’s CPLEX.