2008 IEEE International Symposium on Industrial Electronics, 2008
Final quality of every prediction model is significantly dependent on the quality and good prepro... more Final quality of every prediction model is significantly dependent on the quality and good preprocessing of its input data. A case-based reasoning method is presented for outside temperature data preprocessing for the purpose of short-term gas consumption forecasting. The great advantage of our approach is its simplicity, stability and predictability of results. The suggested method is used both for handling missing data and for user input preprocessing when hourly temperatures profile is computed from few input values. The presented algorithm is computationally robust and is implemented in several real-time systems.
In this paper, the authors describe a system for on-line prediction of truck parking demand along... more In this paper, the authors describe a system for on-line prediction of truck parking demand along a highway system in the Czech Republic. They describe the structure of the system developed during the TACR TA02031411 project and mention some of its specific functionalities. Further, the authors explain in detail statistical modeling methodology which underlies the forecasting model in the core of the prediction procedure. The whole system relies on the use of indirect but very precise and relatively cheap to obtain toll transaction data (accessible through a cooperation with Kapsch Telematic Services, Inc.). The statistical modeling starts with a recognition of the fact that the number of trucks parking at a given lot and given time is a latent variable to be estimated from the observable toll transaction data (which is available in the form of times when individual truck pass toll gates). After constructing an appropriate proxy variable, the authors formulate a flexible class of st...
Energy concept for family house power supply using photovoltaic power plant (PV) with possibility... more Energy concept for family house power supply using photovoltaic power plant (PV) with possibility to operate in off-grid mode was analyzed in this paper. We present partial results from analysis of power consumption of selected house. These results were used as input data for energy concept evaluation and compared with potential energy production calculated from meteorological and climatological data gained from meteorological satellite data.
We investigate various problems encountered when forecasting wind speeds for a network of measure... more We investigate various problems encountered when forecasting wind speeds for a network of measurements stations using outputs of numerical weather prediction (NWP) model as one of the predictors in a statistical forecasting model. First, it is interesting to analyze prediction error properties for different station types (professional and amateur). Secondly, the statistical model can be viewed as a calibration of the original NWP model. Hence, careful semi-parametric smoothing of NWP input can discover various weak points of the NWP, and at the same time, it improves forecasting performance. It turns out that useful information is contained not only in the latest prediction available. It is beneficial to combine different horizon NWP predictions to one target time. GARCH sub-model for the residuals then shows complicated structure usable for short-term forecasts.
We first analyze some features of numerical weather predictions (NWP) for global solar radiation ... more We first analyze some features of numerical weather predictions (NWP) for global solar radiation and notice that they are undersmooth. This finding opens a way to improvements via various smoothing strategies. Then we introduce a statistical modeling framework based on modern semiparametric regression. We use a numerical weather prediction (NWP) model output as one of the inputs for our statistical model. The statistical model is build on the modern regression formalism, utilizing nonparametric B-splines for nonlinear parts whose exact shape is unknown a priori (apart from physically motivated smoothness). Then we illustrate its abilities for systematic development of strategies for NWP calibration and further development. The results are useful both for practical forecasting and as a source of feedback for NWP modelers.
2010 9th International Conference on Environment and Electrical Engineering, 2010
This paper presents a simple and robust wind power forecasting approach using inputs from a state... more This paper presents a simple and robust wind power forecasting approach using inputs from a state-of-the-art numerical weather prediction models (NWP) with mesoscale resolution. The model can be used for short-term and longer term forecasting horizon up to 72 hours ahead. The forecasting ability of the presented approach is demonstrated using real power production data from the Czech Republic.
711 The spatial distribution of air pollution on the local scale of parts of the territory in Cze... more 711 The spatial distribution of air pollution on the local scale of parts of the territory in Czech Republic is simulated by means of Charles University Lagrangian puff model SMOG nested in NWP model ETA. The results are used for the assessment of the concentration fields of ozone, nitrogen oxides and other ozone precursors. A current improved version of the model based on Bednar et al. (2001) covers up to 18 groups of basic compounds and it is based on trajectory computation and puff interaction both by means of Gaussian diffusion, mixing and chemical reactions of basic species. Results of summer photochemical smog episode simulations are compared to results obtained by another couple adopted in the framework of the national project as a basis for further development of data assimilation techniques, Eulerian CTM CAMx nested in NWP model MM5. There are measured data from field campaigns for some episodes as well as air-quality monitoring station data available for comparison of model results with reality. Usually, there is a problem with emission data for the simulations and definitely they are far from actual instantaneous data. Both the couples have rather older databases of emissions available with many uncertainities, for
2014 International Joint Conference on Neural Networks (IJCNN), 2014
Accurate forecasts of weather conditions are of the utmost importance for the management and oper... more Accurate forecasts of weather conditions are of the utmost importance for the management and operation of renewable energy sources with intermittent (stochastic) production. With the growing amount of intermittent energy sources, the need for precise weather predictions increases. Production of energy from renewable power sources, such as wind and solar, can be predicted using numerical weather prediction models. These models can provide high-resolution, localized forecast of wind speed and solar irradiation. However, different instances of numerical weather prediction models may provide different forecasts, depending on their properties and parameterizations. To alleviate this problem, it is possible to employ multiple models and to combine their outputs to obtain more accurate localized forecasts. This work uses the machine-learning tool of Support Vector Regression to amalgamate downward short-wave radiation forecasts of several numerical weather prediction models. Results of SVR-based multi-model forecasts of irradiation at a large set of locations show a significant improvement of prediction accuracy.
2010 IEEE Electrical Power & Energy Conference, 2010
This paper introduces several alternative statistical approaches to modeling and prediction of el... more This paper introduces several alternative statistical approaches to modeling and prediction of electric energy generated by photovoltaic farms. The statistical models use outputs of a numerical weather prediction model as their inputs. Presented statistical models allow for easy-to-compute predictions, both in temporal sense and for out-of-sample individual farms. Model performance is illustrated on a sample of real photovoltaic farms located
ABSTRACT A method is presented for estimating the climatological variability of yearly and monthl... more ABSTRACT A method is presented for estimating the climatological variability of yearly and monthly photovoltaic power production per 1 kWp of installed power. This quantity is computed for a specified portfolio of sources on the basis of historical data. Its climatological variability is derived from a simulation of 33 years of power production with hourly time step. Underlying meteorological variables are taken from the MERRA reanalysis for the years 1979–2011. Since the MERRA reanalysis is not a traditional data source for photovoltaic power modelling, various comparisons to available and more frequently used data sources are included. The method of estimation has the advantage of wide applicability due to the global coverage of the meteorological data.
Quarterly Journal of the Royal Meteorological Society, 2005
An air-quality forecasting system based on the pair 'NWP model MM5-chemistry transport model CAMx... more An air-quality forecasting system based on the pair 'NWP model MM5-chemistry transport model CAMx' is proposed. A version of the ensemble Kalman Filter has been developed. The model-error covariance matrix is parametrized with the help of a covariance function and represented by an ensemble formed as a random selection from leading eigenvectors. The performance of the system is tested on the case of an ozone episode in June 2001. As a source of observations, the AirBase database has been used. Starting the forecast from analysed concentration fields improves the quality of forecast of the next day's ozone concentration maxima.
Novel statistical approaches to prediction have recently been shown to perform well in several sc... more Novel statistical approaches to prediction have recently been shown to perform well in several scientific fields but have not, until now, been comprehensively evaluated for predicting air pollution. In this paper we report on a model inter-comparison exercise in which 15 different statistical techniques for ozone forecasting were applied to ten data sets representing different meteorological and emission conditions throughout Europe. We also attempt to compare the performance of the statistical techniques with a deterministic chemical trajectory model. Likewise, our exercise includes comparisons of sites, performance indices, forecasting horizons, etc. The comparative evaluation of forecasting performance (benchmarking) produced 1340 yearly time series of daily predictions and the results are described in terms of predefined performance indices. Through analysing associations between the performance indices, we found that the success index is of outstanding significance. For models that are excellent in predicting threshold exceedances and have a high success index, we also observe high performance in the overall goodness of fit. The 8-h average ozone concentration forecast accuracy was found to be superior to the 1-h mean ozone concentration forecast, which makes the former very significant for operational forecasting. The best forecasts were achieved for sites located in rural and suburban areas in Central Europe unaffected by extreme emissions (e.g. from industries). Our results demonstrate that a particular technique is often excellent in some respects but poor in others. For most situations, we recommend neural network and generalised additive models as the best compromise, as these can handle nonlinear associations and can be easily adapted to site specific conditions. In contrast, nonlinear modelling of the dynamical development of univariate ozone time-series was not profitable.
Presentation du systeme ELVIRA de prevision de la consommation de gaz, utilise dans la region de ... more Presentation du systeme ELVIRA de prevision de la consommation de gaz, utilise dans la region de Prague. Etude comparative des previsions realisees sur un an, de mars 1999 a mars 2000, avec le systeme ELVIRA et deux methodes, utilisant, pour l'une, des techniques de regression lineaire et pour l'autre, un reseau neuronal.
The authors are engaged in the matematical modelling of gas consumption by small customers and th... more The authors are engaged in the matematical modelling of gas consumption by small customers and they consider an application of one singe general model for the solution to different projects. They describe the GAMMA model in the several different types of tasks, where it was involved. They evaluate this model regarding its possible, more general using.
This tutorial is based on modification of the professor nomination lecture presented two years ag... more This tutorial is based on modification of the professor nomination lecture presented two years ago in front of the Scientific Council of the Czech Technical University in Prague [16]. It is devoted to the techniques for the models developing suitable for processes forecasting in complex systems. Because of the high sensitivity of the processes to the initial conditions and, consequently, due to our limited possibilities to forecast the processes for the long-term horizon, the attention is focused on the techniques leading to practical applications of the short term prediction models. The aim of this tutorial paper is to bring attention to possible difficulties which designers of the predicting models and their users meet and which have to be solved during the prediction model developing, validation, testing, and applications. The presented overview is not complete, it only reflects the author's experience with developing of the prediction models for practical tasks solving in banking, meteorology, air pollution and energy sector. The paper is completed by an example of the global solar radiation prediction which forms an important input for the electrical energy production forecast from renewable sources. The global solar radiation forecasting is based on numerical weather prediction models. The time-lagged ensemble technique for uncertainty quantification is demonstrated on a simple example.
Forecasting procedures are needed when there is uncertainty about the future. In our contribution... more Forecasting procedures are needed when there is uncertainty about the future. In our contribution we discuss some principles that can help to make more accurate forecasts and help to better assess the uncertainty associated with forecasts. We mainly discuss statistical forecasting procedures, but other principles based on experts (judgmental forecasting) and integrating and combining approaches are also mentioned. We show
Applications of Artificial Neural Networks V, 1994
ABSTRACT In this contribution the special interest is focused on the prediction error distributio... more ABSTRACT In this contribution the special interest is focused on the prediction error distributions using feedforward multilayer neural network predictors. The procedure of prediction can be evaluated not only with respect to the mean absolute (or mean square) error but the frequency of higher errors are also taken into account. Two new modifications of the learning algorithms are suggested. The first one is based on error controlled input sample selection for efficient training. The second one is based on the minimization of the special criterial functions, which more reflect especially the great deviations. The functions are expressed in the form of minimax criterial function or in the form of weighted sum of higher order deviations between predicted and measured values. The classical backpropagation is used in the first case and the stochastical method of the statistical gradient is used in the second case. The efficiency of our approaches is demonstrated on the electric load forecasting problem in the west bohemian region in the Czech Republic. A reduction in frequency of higher errors in the everyday morning peaks forecasting were achieved.
2008 IEEE International Symposium on Industrial Electronics, 2008
Final quality of every prediction model is significantly dependent on the quality and good prepro... more Final quality of every prediction model is significantly dependent on the quality and good preprocessing of its input data. A case-based reasoning method is presented for outside temperature data preprocessing for the purpose of short-term gas consumption forecasting. The great advantage of our approach is its simplicity, stability and predictability of results. The suggested method is used both for handling missing data and for user input preprocessing when hourly temperatures profile is computed from few input values. The presented algorithm is computationally robust and is implemented in several real-time systems.
In this paper, the authors describe a system for on-line prediction of truck parking demand along... more In this paper, the authors describe a system for on-line prediction of truck parking demand along a highway system in the Czech Republic. They describe the structure of the system developed during the TACR TA02031411 project and mention some of its specific functionalities. Further, the authors explain in detail statistical modeling methodology which underlies the forecasting model in the core of the prediction procedure. The whole system relies on the use of indirect but very precise and relatively cheap to obtain toll transaction data (accessible through a cooperation with Kapsch Telematic Services, Inc.). The statistical modeling starts with a recognition of the fact that the number of trucks parking at a given lot and given time is a latent variable to be estimated from the observable toll transaction data (which is available in the form of times when individual truck pass toll gates). After constructing an appropriate proxy variable, the authors formulate a flexible class of st...
Energy concept for family house power supply using photovoltaic power plant (PV) with possibility... more Energy concept for family house power supply using photovoltaic power plant (PV) with possibility to operate in off-grid mode was analyzed in this paper. We present partial results from analysis of power consumption of selected house. These results were used as input data for energy concept evaluation and compared with potential energy production calculated from meteorological and climatological data gained from meteorological satellite data.
We investigate various problems encountered when forecasting wind speeds for a network of measure... more We investigate various problems encountered when forecasting wind speeds for a network of measurements stations using outputs of numerical weather prediction (NWP) model as one of the predictors in a statistical forecasting model. First, it is interesting to analyze prediction error properties for different station types (professional and amateur). Secondly, the statistical model can be viewed as a calibration of the original NWP model. Hence, careful semi-parametric smoothing of NWP input can discover various weak points of the NWP, and at the same time, it improves forecasting performance. It turns out that useful information is contained not only in the latest prediction available. It is beneficial to combine different horizon NWP predictions to one target time. GARCH sub-model for the residuals then shows complicated structure usable for short-term forecasts.
We first analyze some features of numerical weather predictions (NWP) for global solar radiation ... more We first analyze some features of numerical weather predictions (NWP) for global solar radiation and notice that they are undersmooth. This finding opens a way to improvements via various smoothing strategies. Then we introduce a statistical modeling framework based on modern semiparametric regression. We use a numerical weather prediction (NWP) model output as one of the inputs for our statistical model. The statistical model is build on the modern regression formalism, utilizing nonparametric B-splines for nonlinear parts whose exact shape is unknown a priori (apart from physically motivated smoothness). Then we illustrate its abilities for systematic development of strategies for NWP calibration and further development. The results are useful both for practical forecasting and as a source of feedback for NWP modelers.
2010 9th International Conference on Environment and Electrical Engineering, 2010
This paper presents a simple and robust wind power forecasting approach using inputs from a state... more This paper presents a simple and robust wind power forecasting approach using inputs from a state-of-the-art numerical weather prediction models (NWP) with mesoscale resolution. The model can be used for short-term and longer term forecasting horizon up to 72 hours ahead. The forecasting ability of the presented approach is demonstrated using real power production data from the Czech Republic.
711 The spatial distribution of air pollution on the local scale of parts of the territory in Cze... more 711 The spatial distribution of air pollution on the local scale of parts of the territory in Czech Republic is simulated by means of Charles University Lagrangian puff model SMOG nested in NWP model ETA. The results are used for the assessment of the concentration fields of ozone, nitrogen oxides and other ozone precursors. A current improved version of the model based on Bednar et al. (2001) covers up to 18 groups of basic compounds and it is based on trajectory computation and puff interaction both by means of Gaussian diffusion, mixing and chemical reactions of basic species. Results of summer photochemical smog episode simulations are compared to results obtained by another couple adopted in the framework of the national project as a basis for further development of data assimilation techniques, Eulerian CTM CAMx nested in NWP model MM5. There are measured data from field campaigns for some episodes as well as air-quality monitoring station data available for comparison of model results with reality. Usually, there is a problem with emission data for the simulations and definitely they are far from actual instantaneous data. Both the couples have rather older databases of emissions available with many uncertainities, for
2014 International Joint Conference on Neural Networks (IJCNN), 2014
Accurate forecasts of weather conditions are of the utmost importance for the management and oper... more Accurate forecasts of weather conditions are of the utmost importance for the management and operation of renewable energy sources with intermittent (stochastic) production. With the growing amount of intermittent energy sources, the need for precise weather predictions increases. Production of energy from renewable power sources, such as wind and solar, can be predicted using numerical weather prediction models. These models can provide high-resolution, localized forecast of wind speed and solar irradiation. However, different instances of numerical weather prediction models may provide different forecasts, depending on their properties and parameterizations. To alleviate this problem, it is possible to employ multiple models and to combine their outputs to obtain more accurate localized forecasts. This work uses the machine-learning tool of Support Vector Regression to amalgamate downward short-wave radiation forecasts of several numerical weather prediction models. Results of SVR-based multi-model forecasts of irradiation at a large set of locations show a significant improvement of prediction accuracy.
2010 IEEE Electrical Power & Energy Conference, 2010
This paper introduces several alternative statistical approaches to modeling and prediction of el... more This paper introduces several alternative statistical approaches to modeling and prediction of electric energy generated by photovoltaic farms. The statistical models use outputs of a numerical weather prediction model as their inputs. Presented statistical models allow for easy-to-compute predictions, both in temporal sense and for out-of-sample individual farms. Model performance is illustrated on a sample of real photovoltaic farms located
ABSTRACT A method is presented for estimating the climatological variability of yearly and monthl... more ABSTRACT A method is presented for estimating the climatological variability of yearly and monthly photovoltaic power production per 1 kWp of installed power. This quantity is computed for a specified portfolio of sources on the basis of historical data. Its climatological variability is derived from a simulation of 33 years of power production with hourly time step. Underlying meteorological variables are taken from the MERRA reanalysis for the years 1979–2011. Since the MERRA reanalysis is not a traditional data source for photovoltaic power modelling, various comparisons to available and more frequently used data sources are included. The method of estimation has the advantage of wide applicability due to the global coverage of the meteorological data.
Quarterly Journal of the Royal Meteorological Society, 2005
An air-quality forecasting system based on the pair 'NWP model MM5-chemistry transport model CAMx... more An air-quality forecasting system based on the pair 'NWP model MM5-chemistry transport model CAMx' is proposed. A version of the ensemble Kalman Filter has been developed. The model-error covariance matrix is parametrized with the help of a covariance function and represented by an ensemble formed as a random selection from leading eigenvectors. The performance of the system is tested on the case of an ozone episode in June 2001. As a source of observations, the AirBase database has been used. Starting the forecast from analysed concentration fields improves the quality of forecast of the next day's ozone concentration maxima.
Novel statistical approaches to prediction have recently been shown to perform well in several sc... more Novel statistical approaches to prediction have recently been shown to perform well in several scientific fields but have not, until now, been comprehensively evaluated for predicting air pollution. In this paper we report on a model inter-comparison exercise in which 15 different statistical techniques for ozone forecasting were applied to ten data sets representing different meteorological and emission conditions throughout Europe. We also attempt to compare the performance of the statistical techniques with a deterministic chemical trajectory model. Likewise, our exercise includes comparisons of sites, performance indices, forecasting horizons, etc. The comparative evaluation of forecasting performance (benchmarking) produced 1340 yearly time series of daily predictions and the results are described in terms of predefined performance indices. Through analysing associations between the performance indices, we found that the success index is of outstanding significance. For models that are excellent in predicting threshold exceedances and have a high success index, we also observe high performance in the overall goodness of fit. The 8-h average ozone concentration forecast accuracy was found to be superior to the 1-h mean ozone concentration forecast, which makes the former very significant for operational forecasting. The best forecasts were achieved for sites located in rural and suburban areas in Central Europe unaffected by extreme emissions (e.g. from industries). Our results demonstrate that a particular technique is often excellent in some respects but poor in others. For most situations, we recommend neural network and generalised additive models as the best compromise, as these can handle nonlinear associations and can be easily adapted to site specific conditions. In contrast, nonlinear modelling of the dynamical development of univariate ozone time-series was not profitable.
Presentation du systeme ELVIRA de prevision de la consommation de gaz, utilise dans la region de ... more Presentation du systeme ELVIRA de prevision de la consommation de gaz, utilise dans la region de Prague. Etude comparative des previsions realisees sur un an, de mars 1999 a mars 2000, avec le systeme ELVIRA et deux methodes, utilisant, pour l'une, des techniques de regression lineaire et pour l'autre, un reseau neuronal.
The authors are engaged in the matematical modelling of gas consumption by small customers and th... more The authors are engaged in the matematical modelling of gas consumption by small customers and they consider an application of one singe general model for the solution to different projects. They describe the GAMMA model in the several different types of tasks, where it was involved. They evaluate this model regarding its possible, more general using.
This tutorial is based on modification of the professor nomination lecture presented two years ag... more This tutorial is based on modification of the professor nomination lecture presented two years ago in front of the Scientific Council of the Czech Technical University in Prague [16]. It is devoted to the techniques for the models developing suitable for processes forecasting in complex systems. Because of the high sensitivity of the processes to the initial conditions and, consequently, due to our limited possibilities to forecast the processes for the long-term horizon, the attention is focused on the techniques leading to practical applications of the short term prediction models. The aim of this tutorial paper is to bring attention to possible difficulties which designers of the predicting models and their users meet and which have to be solved during the prediction model developing, validation, testing, and applications. The presented overview is not complete, it only reflects the author's experience with developing of the prediction models for practical tasks solving in banking, meteorology, air pollution and energy sector. The paper is completed by an example of the global solar radiation prediction which forms an important input for the electrical energy production forecast from renewable sources. The global solar radiation forecasting is based on numerical weather prediction models. The time-lagged ensemble technique for uncertainty quantification is demonstrated on a simple example.
Forecasting procedures are needed when there is uncertainty about the future. In our contribution... more Forecasting procedures are needed when there is uncertainty about the future. In our contribution we discuss some principles that can help to make more accurate forecasts and help to better assess the uncertainty associated with forecasts. We mainly discuss statistical forecasting procedures, but other principles based on experts (judgmental forecasting) and integrating and combining approaches are also mentioned. We show
Applications of Artificial Neural Networks V, 1994
ABSTRACT In this contribution the special interest is focused on the prediction error distributio... more ABSTRACT In this contribution the special interest is focused on the prediction error distributions using feedforward multilayer neural network predictors. The procedure of prediction can be evaluated not only with respect to the mean absolute (or mean square) error but the frequency of higher errors are also taken into account. Two new modifications of the learning algorithms are suggested. The first one is based on error controlled input sample selection for efficient training. The second one is based on the minimization of the special criterial functions, which more reflect especially the great deviations. The functions are expressed in the form of minimax criterial function or in the form of weighted sum of higher order deviations between predicted and measured values. The classical backpropagation is used in the first case and the stochastical method of the statistical gradient is used in the second case. The efficiency of our approaches is demonstrated on the electric load forecasting problem in the west bohemian region in the Czech Republic. A reduction in frequency of higher errors in the everyday morning peaks forecasting were achieved.
Uploads
Papers by Emil Pelikan