a b s t r a c tLarge amounts of new wind power are currently under construction or planning in many countries. The constantly increasing percentage of wind power in the electricity generation mix has to be taken into consideration when planning power systems. This paper introduces a Monte Carlo simulation based methodology that can be used to assess the effects (e.g. need for new transmission lines, reserves, wind curtailment or demand side management) of large amounts of existing and planned wind power generation on the power system. The presented methodology is able to assess new wind power scenarios spread over a wide geographical area, comprising numerous existing and planned wind generation locations. The Monte Carlo simulation results are verified against measured aggregated wind power generation in Finland from 2008 to 2014. In addition, case studies of future scenarios with 232 individual wind generation locations are presented to show the applicability of the methodology as a tool in power system planning.

Improved performance electricity demand forecast can provide decentralized energy system operators, aggregators, managers, and other stakeholders with essential information for energy resource scheduling, demand response management, and energy market participation. Most previous methodologies have focused on predicting the aggregate amount of electricity demand at national or regional scale and disregarded the electricity demand for small-scale decentralized energy systems (buildings, energy communities, microgrids, local energy internets, etc.), which are emerging in the smart grid context. Furthermore, few research groups have performed attribute selection before training predictive models. This paper proposes a machine learning (ML)-based integrated feature selection approach to obtain the most relevant and nonredundant predictors for accurate short-term electricity demand forecasting in distributed energy systems. In the proposed approach, one of the ML tools -binary genetic algorithm (BGA) is applied for the feature selection process and Gaussian process regression (GPR) is used for measuring the fitness score of the features. In order to validate the effectiveness of the proposed approach, it is applied to various building energy systems located in the Otaniemi area of Espoo, Finland. The findings are compared with those achieved by other feature selection techniques. The proposed approach enhances the quality and efficiency of the predictor selection, with minimal chosen predictors to achieve improved prediction accuracy. It outperforms the other evaluated feature selection methods. Besides, a feedforward artificial neural network (FFANN) model is implemented to evaluate the forecast performance of the selected predictor subset. The model is trained using two-year hourly dataset and tested with another one-year hourly dataset. The obtained results verify that the FFANN forecast model based on the BGA-GPR FS selected training feature subset has achieved an annual MAPE of 1.96%, which is a very acceptable and promising value for electricity demand forecasting in small-scale decentralized energy systems.INDEX TERMS Binary genetic algorithm, decentralized energy system, electricity demand forecasting, feature selection, feedforward artificial neural network, fitness evaluation measure, Gaussian process regression, machine learning, smart grid.

As installed wind generation capacity increases, understanding the effect of wind power on the electric power system is becoming more important. This paper introduces a statistical model that can be used to estimate the variability in wind generation and assess the risk of wind generation contingencies over a large geographical area. The analysis of the installed wind generation capacities is separated from the analysis of the spatial and temporal dependency structures. This enables the study of different future wind power scenarios with varying generation capacities. The model is built on measured hourly wind generation data from Denmark, Estonia, Finland and Sweden. Three scenarios with different geographical distributions of wind power are compared to show the applicability of the model for power system planning. A method for finding the scenario with the minimum variance of the aggregate wind generation is introduced. As the geographical distribution of wind power can be affected by subsidies and other incentives, the presented results can have policy implications.

The judicious placement of disconnecting switches is an efficient means to enhance the reliability of distribution networks. Aiming at optimizing the investment in these switches, this paper presents a mathematical programming-based model considering the installation of remote-controlled and manual switches at various locations in the distribution network. The proposed model not only yields the optimal location and type of switches in the main feeders but also specifies the optimal type of tie switches, i.e., backup switches at the reserve connection points. Incentive reliability regulation in the form of a reward-penalty scheme is incorporated into the proposed model to take the distribution service reliability worth into account realistically. In addition to this cost, the revenue lost due to energy undelivered during the distribution network faults is considered to determine the unreliability costs more accurately. In order to estimate such reliability-related costs, a novel reliability assessment technique is developed and integrated into the proposed switch optimization model. Formulated as an instance of mixed-integer linear programming, the proposed model is applied to a test distribution network, and the outcomes are investigated in detail. INDEX TERMS Electricity distribution system, mixed-integer linear programming, reliability, reward-penalty scheme, switch optimization. NOMENCLATURE INDICES

Abstract-The analysis of large-scale wind and photovoltaic (PV) energy generation is of vital importance in power systems where their penetration is high. This paper presents a modular methodology to assess the power generation and volatility of a system consisting of both PV plants (PVPs) and wind power plants (WPPs) in new locations. The methodology is based on statistical modelling of PV and WPP locations with a vector autoregressive model, which takes into account both the temporal correlations in individual plants and the spatial correlations between the plants. The spatial correlations are linked through distances between the locations, which allows the methodology to be used to assess scenarios with PVPs and WPPs in multiple locations without actual measurement data. The methodology can be applied by the transmission and distribution system operators when analysing the effects and feasibility of new PVPs and WPPs in system planning. The model is verified against hourly measured wind speed and solar irradiance data from Finland. A case study assessing the impact of the geographical distribution of the PVPs and WPPs on aggregate power generation and its variability is presented.