The short term forecasting of wind speed and direction has previously been improved by adopting a cyclo-stationary multichannel linear prediction approach which incorporated seasonal cycles into the estimation of statistics. This paper expands previous analysis by also incorporating diurnal variation and time-dependent window lengths. Based on a large data set from the UK’s Met Office, we demonstrate the impact of this proposed approach.

A spatio-temporal method for producing very-short-term parametric probabilistic wind power forecasts at a large number of locations is presented. Smart grids containing tens, or hundreds, of wind generators require skilled very-short-term forecasts to operate effectively and spatial information is highly desirable. In addition, probabilistic forecasts are widely regarded as necessary for optimal power system management as they quantify the uncertainty associated with point forecasts. Here, we work within a parametric framework based on the logit-normal distribution and forecast its parameters. The location parameter for multiple wind farms is modeled as a vector-valued spatio-temporal process and the scale parameter is tracked by modified exponential smoothing. A state-of-the-art technique for fitting sparse vector autoregressive models is employed to model the location parameter and demonstrates numerical advantages over conventional vector autoregressive models. The proposed method is tested on a dataset of 5 min mean wind power generation at 22 wind farms in Australia. Five-min ahead forecasts are produced and evaluated in terms of point, and probabilistic forecast skill scores and calibration. Conventional autoregressive and vector autoregressive models serve as benchmarks.

In this paper we examine potential improvements in how load and generation forecast uncertainty is captured when setting reserve levels in power systems with significant renewable generation penetration and discuss the merit of proposed new methods in this area.One important difference between methods is whether reserves are defined based on the marginal distribution of forecast errors, as calculated from historic data, or whether the conditional distribution, specific to the time at which reserves are being scheduled, is used.This paper is a review of published current practice in markets which are at the leading edge of this problem, summarizing their experiences, and aligning it with academic modeling work.We conclude that the ultimate goal for all markets expected to manage high levels of renewable generation should be a reserve setting mechanism which utilizes the best understanding of meteorological uncertainties combined with traditional models of uncertainty arising from forced outages.

AbstractPresently, in several countries, the wind and solar power capacity connected to the distribution network is increasing steadily. In parallel, the investment in smart grid technologies is growing, which will support the integration of renewable energy and represents an opportunity to develop advanced management functions. Forecasting is a key input for several management functions of distribution system operators (DSOs).This chapter describes modeling approaches allowing to take advantage of the wealth of power measurements available in near real‐time in a smart grid context [e.g., smart meters and SCADA (supervisory control and data acquisition) system], in order to improve the quality of renewable energy forecasting in a computationally efficient manner. As a basis, the power generation vector is to be considered as a multivariate one, that is, by simultaneously focusing on all sites of interest, instead of trying to build and estimate models at every location, individually. Two real‐world test cases with wind and solar generation are used to show the improvement in accuracy from exploring distributed information in a probabilistic forecasting framework.

Two nonlinear methods for producing short-term spatio-temporal wind speed forecast are presented. From the relatively new class of kernel methods, a kernel least mean squares algorithm and kernel recursive least squares algorithm are introduced and used to produce 1 to 6 hour-ahead predictions of wind speed at six locations in the Netherlands. The performance of the proposed methods are compared to their linear equivalents, as well as the autoregressive, vector autoregressive and persistence time series models. The kernel recursive least squares algorithm is shown to offer significant improvement over all benchmarks, particularly for longer forecast horizons. Both proposed algorithms exhibit desirable numerical properties and are ripe for further development.

This paper demonstrates that wave height forecasters chosen on statistical quality metrics result in sub-optimal decision support for offshore wind farm maintenance. Offshore access is constrained by wave height, but the majority of approaches to evaluating the effectiveness of a wave height forecaster utilize overall accuracy or error rates. This paper introduces a new metric more appropriate to the wind industry, which considers the economic impact of an incorrect forecast above or below critical wave height boundaries. The paper describes a process for constructing a value criteria where the implications between forecasting error and economic consequences are explicated in terms of opportunity costs and realized maintenance costs. A comparison between nine forecasting techniques for modeling and predicting wave heights based on historical data, including an ensemble aggregator, is described demonstrating that the performance ranking of forecasters is sensitive to the evaluation criteria. The results highlight the importance of appropriate metrics for wave height prediction specific to the wind industry, and the limitations of current models that minimize a metric that does not support decision making. With improved ability to forecast weather windows, maintenance scheduling is subject to less uncertainty, hence reducing costs related to vessel dispatch, and lost energy due to downtime.

The short term forecasting of wind speed and direction has previously been improved by adopting a cyclo-stationary multichannel linear prediction approach which incorporated seasonal cycles into the estimation of statistics.This paper expands previous analysis by also incorporating diurnal variation and time-dependent window lengths.Based on a large data set from the UK's Met Office, we demonstrate the impact of this proposed approach.

The desire to improve short-term predictions of wind speed and direction has motivated the development of a spatial covariance-based predictor in a complex valued multichannel structure. Wind speed and direction are modelled as the magnitude and phase of complex time series and measurements from multiple geographic locations are embedded in a complex vector which is then used as input to a multichannel Wiener prediction filter. Building on a C-linear cyclo-stationary predictor, a new widely linear filter is developed and tested on hourly mean wind speed and direction measurements made at 13 locations in the UK over 6 years. The new predictor shows a reduction in mean squared error at all locations. Furthermore it is found that the scale of that reduction strongly depends on conditions local to the measurement site.

This paper aims to produce a low-complexity predictor for the hourly mean wind speed and direction from 1 to 6h ahead at multiple sites distributed around the UK. The wind speed and direction are modelled via the magnitude and phase of a complex-valued time series. A multichannel adaptive filter is set to predict this signal on the basis of its past values and the spatio-temporal correlation between wind signals measured at numerous geographical locations. The filter coefficients are determined by minimizing the mean square prediction error. To account for the time-varying nature of the wind data and the underlying system, we propose a cyclo-stationary Wiener solution, which is shown to produce an accurate predictor. An iterative solution, which provides lower computational complexity, increased robustness towards ill-conditioning of the data covariance matrices and the ability to track time-variations in the underlying system, is also presented. The approaches are tested on wind speed and direction data measured at various sites across the UK. Results show that the proposed techniques are able to predict wind speed as accurately as state-of-the-art wind speed forecasting benchmarks while simultaneously providing valuable directional information. Copyright (c) 2013 John Wiley & Sons, Ltd.

This paper discusses analysis of wind and wave data to assess maintenance access for offshore wind farms.

This paper proposes a statistical method for 1-6 hour-ahead prediction of hourly mean wind speed and direction to better forecast the power produced by wind turbines, an increasingly important component of power system operation. The wind speed and direction are modelled via the magnitude and phase of a complex vector containing measurements from multiple geographic locations. The predictor is derived from the spatio-temporal covariance which is estimated at regular time intervals from a subset of the available training data, the wind direction of which lies within a sliding range of angles centred on the most recent measurement of wind direction. This is a generalisation of regime-switching type approaches which train separate predictors for a few fixed regimes. The new predictor is tested on the Hydra dataset of wind across the Netherlands and compared to persistence and a cyclo-stationary Wiener filter, a state-of-the-art spatial predictor of wind speed and direction. Results show that the proposed technique is able to predict the wind vector more accurately than these benchmarks on dataset containing 4 to 27 sites, with greater accuracy for larger datasets.

Due to the large and increasing penetration of wind power around the world, accurate power production forecasts are required to manage power systems and wind power plants. In this paper we propose an ensemble of particle swarm optimised filtering technique for 1-hour-ahead prediction of hourly mean wind speed and direction. The performance of the new method is assessed by testing it on data from 13 locations around the UK where it performs comparably to linear techniques but is able to provide significant improvement at a subset of locations.

This paper develops a linear predictor for application to wind speed and direction forecasting in time and across different sites. The wind speed and direction are modelled via the magnitude and phase of a complex-valued time-series. A multichannel adaptive filter is set to predict this signal, based on its past values and the spatio-temporal correlation between wind signals measured at numerous geographical locations. The time-varying nature of the underlying system and the annual cycle of seasons motivates the development of a cyclostationary Wiener filter, which is tested on hourly mean wind speed and direction data from 13 weather stations across the UK, and shown to provide an improvement over both stationary Wiener filtering and a recent auto-regressive approach.