This paper describes a complete digitally controlled dc-dc buck converter performed by field-programmable gate array (FPGA) circuitry. The voltage and current-mode control is based on a voltage control oscillator (VCO) performed measurements regarding output-voltage and inductor current, respectively. This measurement principle also uses digital-counters (digital-integrators) in order to obtain integral values for the output-voltage and inductor current. In analog current-mode control, the instantaneous inductor current-value-measurement is used for the switching action. When the VCO is used for the inductor current measurement, the integral is measured during the switching-on time set as an observation interval and the switching action occurs based on this measurement. Such a principle enables full digitalization of the voltage- and current-control loop and also the used measurement principle is capable of rejecting the switching disturbances during current and voltage measurements. All the tasks for current and voltage control were implemented within the FPGA. The algorithm was verified by simulation and experimentation at a switching-frequency of 25kHz.
COBISS.SI-ID: 17671446
One of the major challenges today is assessing the suitability of PV (photovoltaic) systems' installations on buildings' roofs regarding the received solar irradiance. The availability of aerial laser-scanning, namely LiDAR (Light Detection And Ranging), means that assessment can be performed automatically over large-scale urban areas in high accuracy by considering surfaces' topographies, long-term direct and diffuse irradiance measurements, and influences of shadowing. The solar potential metric was introduced for this purpose, however it fails to provide any insights into the production of electrical energy by a specific PV system. Hence, the PV potential metric can be used that integrates received instantaneous irradiance which is then multiplied by the PV system's efficiency characteristics. Many existing PV potential metrics over LiDAR data consider the PV modules' efficiencies to be constant, when in reality they are nonlinear. This paper presents a novel PV potential estimation over LiDAR data, where the PV modules' and solar inverter's nonlinear efficiency characteristics are approximated by modelled functions. The estimated electrical energy production from buildings' roofs within an urban area was extensively analyzed by comparing the constant and nonlinear efficiency characteristics of different PV module types and solar inverters. The obtained results were confirmed through measurements performed on an existing PV system. The efficiency characteristics used in this paper where measured on micro-inverter used in the project as a part of the low voltage network control and as a part of control system that controls individual groups of micro inverters. The placement of PV systems used in the project is determined by the method proposed in this paper, considering also properties of available electricity network.
COBISS.SI-ID: 17543702
The work evaluates the usability of the stochastic direct search algorithm called differential evolution (DE) for the simultaneous identification of all parameters of an electric circuit which is in the give case a special synchrnous machine. The used algorithm has several advantages in determination of the different operating states of electric power networks as well. Namely, it is robust in attaining global minima, it is appropriate for solving nonlinear and constrained optimization problems, it needs only a few control parameters to be defined, and it does not require a starting point, but boundaries of expected solutions. For the aforementioned reasons, the DE procedure is used in activities of the research project as well; to be more precise, in the determination of power flows in a distribution electric network with large share of distributed power sources, as well as in the determination of optimal distribution of reactive and active power generation among individual distributed generation units.
COBISS.SI-ID: 17638166