Marti, P.; Velasco, M.; Torres, J.; Miret, J.; Castilla, M. IEEE International Conference on Control and Automation p. 371-376 DOI: 10.1109/ICCA.2016.7505305 Data de presentació: 2016-06 Presentació treball a congrés
Guzman, R.; Garcia de Vicuña, J.; Morales, J.; Castilla, M.; Miret, J.; Torres, J. Annual Conference of the IEEE Industrial Electronics Society p. 1211-1216 DOI: 10.1109/IECON.2015.7392265 Data de presentació: 2015-11-12 Presentació treball a congrés
This paper presents an improved variable hysteresis-band current-control in natural frame for a three-phase shunt active power filter. The proposed control algorithm is based on three decoupled sliding-mode controllers combined with three independent Kalman filters. The use of Kalman filters instead of a non-adaptive state observer improves the quality of the estimated signals in presence of noise, increasing the immunity of the control loop in noisy environments and also reducing the THD of the current delivered to the grid. The overall control proposal has been fully integrated into a digital signal processor. Selected experimental results are introduced to validate the theoretical contributions of this paper.
Velasco, M.; Marti, P.; Torres, J.; Miret, J.; Castilla, M. Annual Conference of the IEEE Industrial Electronics Society p. 3755-3760 DOI: 10.1109/IECON.2015.7392686 Data de presentació: 2015-11-11 Presentació treball a congrés
The increasing deployment of distributed generation (DG) such as photovoltaic panels (PV) connected to low-voltage (LV) grids is becoming a common trend in urban areas. The advances of information and communication technology (ICT) facilitates the collaborative operation of DG systems to achieve collective benefits. The fusion of these two trends creates a new scenario where reactive power control methods can offer
additional features and benefits beyond the conventional voltage regulation provided by the droop method. Taking advantage of this new scenario, this paper formulates the application of reactive power control as an optimization problem where simple and ideal settings are imposed by design in order to facilitate the exploration search as well as to avoid over-constraining the optimization space. By appropriately using the reactive power capacity of inverters, the desired collective benefit is to minimize
power losses while individual voltages at each inverter should be kept within the statutory limits. The simulated solution of the optimization problem is applied to a real-inspired PV-LV grid subject to an over-voltage situation, which may typically occur during periods of high production but low consumption.
Simulation results reveal unexpected optimal settings for reactive power control set-points at each inverter, which calls for a final discussion to review the applicability of the optimization approach.
Torres, J.; Castilla, M.; Miret, J.; Moradi, M.; Morales, J.; Guzman, R. Annual Conference of the IEEE Industrial Electronics Society p. 1228-1233 DOI: 10.1109/IECON.2015.7392268 Data de presentació: 2015-11-11 Presentació treball a congrés
Three-phase inverters are the mostly used converters for connecting distributed generation sources to the main grid. Conventionally, these renewable sources must inject only its generated active power. Recently, reactive power injection has been included in some national grid codes. In voltage regulation, the conventional PI control can not be used. The whole system tends to instability when PI controllers work distributively for voltage regulation in different connection points along the grid. The slope voltage controller seems to be the best option to avoid these instabilities. The parameters design of the slope voltage controller is normally carried out by trial and error methodology due to lack of a small-signal model of the complete system. With this widely used methodology, the dynamic behavior can not be determined by the designer. This work develops a closed-loop small signal model of a grid-connected distributed generation source with voltage regulation capabilities. The derived mathematical model predicts accurately the system behavior in the frequency range of interest. This model can be used in future research for tuning the controller parameters with the aim of achieve predefined design objectives. The model has been validated through simulation results.
Guzman, R.; Garcia de Vicuña, J.; Morales, J.; Momeneh, A.; Miret, J.; Torres, J. Annual Conference of the IEEE Industrial Electronics Society p. 1217-1222 DOI: 10.1109/IECON.2015.7392266 Data de presentació: 2015-11-10 Presentació treball a congrés
This paper presents an active damping method in natural frame for a three-phase voltage source inverter with LCL filter. The proposed method is based on the pole placement technique via Ackermann's formula. This approach is used to obtain the proper sliding surface vector coefficients to emulate a virtual resistor in series with the capacitor filter. Besides a well-known method in the literature have been used to obtain three decoupled controllers in natural frame. The stability is theoretically studied and experimental results shows the validity of this proposal.
Morales, J.; Garcia de Vicuña, J.; Guzman, R.; Castilla, M.; Miret, J.; Torres, J. Annual Conference of the IEEE Industrial Electronics Society p. 4519-4524 DOI: 10.1109/IECON.2015.7392804 Data de presentació: 2015-11-09 Presentació treball a congrés
This paper present a sliding mode control operating at fixed switching frequency for a three-phase unity power factor rectifier using the vector operation technique. The proposed method allows to find a large-signal converter dynamic model in order to transform a tree-phase system into two decoupled subsystems. With this technique only two sliding-mode controllers are needed and the neutral point voltage influence is removed from the controllers dynamics. A fast output-voltage control is used which avoids output-voltage variations when a sudden change in the load appears. Besides a variable hysteresis band control is used in order to fix the switching frequency of the sliding-mode controllers. Experimental results are provided using a fully-digital control system in order to validate the theoretical predictions.
Morales, J.; Garcia de Vicuña, J.; Guzman, R.; Castilla, M.; Momeneh, A.; Torres, J. Annual Conference of the IEEE Industrial Electronics Society p. 1182-1187 DOI: 10.1109/IECON.2015.7392260 Data de presentació: 2015-11 Presentació treball a congrés
This paper presents a sliding mode control for a three-phase grid-connected voltage source inverter with vector operation. The design of closed-loop system is based on the sliding-mode theory, and it is carried out with the help of a converter large-signal dynamic model. The closed-loop system improves the transient response and minimizes the steady-state error. Essentially, the contribution of this paper is as follows: the dynamic model for the tree-phase grid-connected inverter with vector operation and the proposed sliding-mode control scheme.
Experimental results are provided using a fully-digital control system in order to validate the theoretical predictions.
Marti, P.; Velasco, M.; Martin, E.X.; Castilla, M.; Miret, J.; Torres, J. IEEE International Conference on Industrial Informatics p. 191-196 DOI: 10.1109/INDIN.2015.7281733 Data de presentació: 2015-07-17 Presentació treball a congrés
A topic of interest within the microgrid (MG) community is to achieve accurate sharing of active and reactive power among voltage source inverters (VSI) working in parallel while keeping the voltage frequency and amplitude stable, and close to a given reference. This problem has been mainly solved by the so-called frequency and voltage droop controllers that are local control loops implemented in each VSI. However, since they introduce a frequency and amplitude deviation, a secondary integral-like control loop using a communication system is usually deployed to enable each VSI to cancel the deviations. This paper evaluates two secondary control implementations of the frequency restoration with respect to properties of the communication system. Networked control applications are subject to known problems that are inherent to the use of a communication system such as message losses or varying sampling/transmission intervals, to name a few. The performance analysis reveals that control policies should be designed accounting also for the properties of the communication system. Otherwise, unexpected results violating the control specifications may appear.