Abstract
High penetration of Photovoltaic (PV) systems, a variable resource, poses challenges to the stability and power quality of electrical grids. Forecasting accurately the PV power has been recognized as a way to ease this problem. This work addresses now-casting of PV power with Extreme Learning Machines (ELMs) without exogenous inputs. Wavelet decomposition and multi-resolution analysis is the most effective way to achieve high accuracy for 5 min-ahead forecast up to 70% greater than the persistence model. A neural network evaluation algorithm based on multiple initializations and incremental hidden nodes is applied and ELMs performance and computation efficiency is evaluated versus Time Delay Neural Networks (TDNNs) for time and time-frequency domain forecasting.

Abstract
One of the major changes currently involving distribution network systems (DNSs) is the ever-increasing integration of renewable-based distributed generation (DG), wind and solar PV types in particular. This is dramatically influencing the planning and operation of distribution systems, in general. The traditional "fit-and-forget" approach is outdated. Current developments in the DNS would require new, efficient and robust planning and operation tools to support smooth integration of such DGs. The present work focuses on an optimal integration of renewable-based DGs with reactive power support capabilities. Accordingly, a stochastic mixed integer linear programming (S-MILP) model is developed that takes into account the optimal integration of RES-based DGs and reactive power sources. The developed model is tested using a standard IEEE distribution system. Test results show that integrating DGs with reactive power support capability significantly enhances voltage stability and improves the overall cost in the system. Simulation results show that setting the reactive power support capability of the RES-based DGs from 0.95 leading to 0.95 lagging leads to the maximum penetration level of wind and solar PV power in the system.

Abstract
A control technique is proposed in this paper for control of modular multilevel converters (MMC) in high-voltage direct current (HVDC) transmission systems. Six independent dynamical state variables are considered in the proposed control technique, including two ac currents, three circulating currents, and the dc-link voltage, for effectively attaining the switching state functions of MMCs, as well as for an accurate control of the circulating currents. Several analytical expressions are derived based on the reference values of the state variables for obtaining the MMC switching functions under steady state operating conditions. In addition, dynamic parts of the switching functions are accomplished by the direct Lyapunov method to guarantee stable operation of the proposed technique for control of MMCs in HVDC systems. Moreover, the capability curve of MMC is developed to validate maximum power injection from MMCs into the power grid and/or loads. The impacts of the variations of MMC output and dc-link currents on the stability of dc-link voltage are also evaluated in detail by small-signal analysis.

Abstract
A new multistage and stochastic mathematical model of an integrated distribution system planning problem is described in Part I. The efficiency and validity of this model are tested by carrying out a case study on a standard IEEE 41-bus radial distribution system. The numerical results show that the simultaneous integration of energy storage systems (ESSs) and reactive power sources largely enables a substantially increased penetration of variable generation (wind and solar) in the system, and consequently, reduces overall system costs and network losses. For the system, a combined wind and solar PV power of up to nearly three times the base-case peak load is installed over a three-year planning horizon. In addition, the proposed planning approach also considerably defers network expansion and/or reinforcement needs. Generally, it is clearly demonstrated in an innovative way that the joint planning of distributed generation, reactive power sources, and ESSs, brings significant improvements to the system such as reduction of losses, electricity cost, and emissions as a result of increased renewable energy sources (RESs) penetration. Besides, the proposed modeling framework considerably improves the voltage profile in the system, which is crucial for a normal operation of the system as a whole. Finally, the novel planning model proposed can be considered as a major leap forward toward developing controllable grids, which support large-scale integration of RESs.

Abstract
The integration of storage systems into smart grids is being widely analysed in order to increase the flexibility of the power system and its ability to accommodate a higher share of wind and solar power. The success of this process requires a comprehensive techno-economic study of the storage technology in contrast with electricity market behaviour. The focus of this work is on lead-acid and vanadium redox flow batteries. This paper presents a novel probabilistic optimization model for managing energy storage systems. The model is able to incorporate the forecasting error of electricity prices, offering with this a near-optimal control option. Using real data from the Spanish electricity market from the year 2016, the probability distribution of forecasting error is determined. The model determines electricity price uncertainty by means of Monte Carlo Simulation and includes it in the energy arbitrage problem, which is eventually solved by using an integer-coded genetic algorithm. In this way, the probability distribution of the revenue is determined with consideration of the complex behaviours of lead acid and vanadium redox flow batteries as well as their associated operating devices such as power converters.

Abstract
Due to the uncertainty and stochastic behavior of wind and photovoltaic production introduced in conventional power systems, the correct overall management considering all the technical and economic constraints is faced with more challenges. To address also the specificities of insular power systems, several strategies have been proposed in last years, including energy storage systems with the aim of increasing system flexibility. Accurate forecasting tools may also help to reduce overall uncertainty. Other scheduling tools based on probabilistic, heuristic and stochastic programming have also been considered. In this work, a new scheduling strategy is proposed considering the integration of wind production in an insular power system. To this end, some arbitrarily chosen scenarios from wind production are introduced in the scheduling process, and a comparative study is carried out, with and without renewable production, providing an acceptable computational time.