This paper presents an analysis to forecast the loads of an isolated area where the history of load is not available or the history may not represent the realistic demand of electricity. The analysis is done through l...This paper presents an analysis to forecast the loads of an isolated area where the history of load is not available or the history may not represent the realistic demand of electricity. The analysis is done through linear regression and based on the identification of factors on which electrical load growth depends. To determine the identification factors, areas are selected whose histories of load growth rate known and the load growth deciding factors are similar to those of the isolated area. The proposed analysis is applied to an isolated area of Bangladesh, called Swandip where a past history of electrical load demand is not available and also there is no possibility of connecting the area with the main land grid system.展开更多
The fuzzy linear regression forecasting model is deduced from the symmetric triangular fuzzy number. With the help of the degree of fitting and the measure of fuzziness, the determination of symmetric triangular fuzzy...The fuzzy linear regression forecasting model is deduced from the symmetric triangular fuzzy number. With the help of the degree of fitting and the measure of fuzziness, the determination of symmetric triangular fuzzy numbers is changed into a problem of solving linear programming.展开更多
The present paper discusses the application of localized linear models for the prediction of hourly PM10 concentration values. The advantages of the proposed approach lies in the clustering of the data based on a comm...The present paper discusses the application of localized linear models for the prediction of hourly PM10 concentration values. The advantages of the proposed approach lies in the clustering of the data based on a common property and the utilization of the target variable during this process, which enables the development of more coherent models. Two alternative localized linear modelling approaches are developed and compared against benchmark models, one in which data are clustered based on their spatial proximity on the embedding space and one novel approach in which grouped data are described by the same linear model. Since the target variable is unknown during the prediction stage, a complimentary pattern recognition approach is developed to account for this lack of information. The application of the developed approach on several PM10 data sets from the Greater Athens Area, Helsinki and London monitoring networks returned a significant reduction of the prediction error under all examined metrics against conventional forecasting schemes such as the linear regression and the neural networks.展开更多
时间序列预测在能源管理、交通流量和气象分析等多个实际场景中具有重要应用价值。然而,时间序列数据中存在的分布漂移(Distribution Shift)与长程依赖(Long-term Dependency)仍限制了传统方法与现有深度学习模型在长期预测中的表现。为...时间序列预测在能源管理、交通流量和气象分析等多个实际场景中具有重要应用价值。然而,时间序列数据中存在的分布漂移(Distribution Shift)与长程依赖(Long-term Dependency)仍限制了传统方法与现有深度学习模型在长期预测中的表现。为此,提出了一种名为D-LINet(Dual-Normalization and Linear Integration Network)的创新模型。该模型结合了Dish-TS(Distribution Shift in Time Series Forecasting)框架的分布归一化能力与线性映射的高效性,并采用双向归一化与双线性层的设计,有效缓解输入与输出空间的分布偏移,增强了对周期性与趋势性特征的捕捉能力。在多个真实数据集上对D-LINet的预测性能进行了全面评估。结果显示,在短期与长期预测中,D-LINet的均方误差和平均绝对误差均显著优于主流模型(如Transformer,Informer,Autoformer和DLinear)。此外,实验还探讨了输入窗口长度及先验知识的引入对预测性能的影响,为后续模型优化提供了重要指导。该研究针对复杂分布漂移问题提出了新的解决思路,并有助于提升时间序列预测的精度与稳健性。展开更多
The 3-hour-interval prediction of ground-level temperature from +00 h out to +45 h in South Korea (38 stations) is performed using the DLM (dynamic linear model) in order to eliminate the systematic error of numerical...The 3-hour-interval prediction of ground-level temperature from +00 h out to +45 h in South Korea (38 stations) is performed using the DLM (dynamic linear model) in order to eliminate the systematic error of numerical model forecasts. Numerical model forecasts and observations are used as input values of the DLM. According to the comparison of the DLM forecasts to the KFM (Kalman filter model) forecasts with RMSE and bias, the DLM is useful to improve the accuracy of prediction.展开更多
One of the principal issues related to hydrologic models for prediction of runoff is the estimation of extreme values (floods). It is well understood that unless the models capture the dynamics of rainfall-runoff proc...One of the principal issues related to hydrologic models for prediction of runoff is the estimation of extreme values (floods). It is well understood that unless the models capture the dynamics of rainfall-runoff process, the improvement in prediction of such extremes is far from reality. In this paper, it is proposed to develop a dual (combined and paralleled) artificial neural network (D-ANN), which aims to improve the models performance, especially in terms of extreme values. The performance of the proposed dual-ANN model is compared with that of feed forward ANN (FF-ANN) model, the later being the most common ANN model used in hydrologic literature. The forecasting exercise is carried out for hourly river flow data of Kolar Basin, India. The results of the comparison indicate that the D-ANN model performs better than the FF-ANN model.展开更多
In this paper, the model output machine learning (MOML) method is proposed for simulating weather consultation, which can improve the forecast results of numerical weather prediction (NWP). During weather consultation...In this paper, the model output machine learning (MOML) method is proposed for simulating weather consultation, which can improve the forecast results of numerical weather prediction (NWP). During weather consultation, the forecasters obtain the final results by combining the observations with the NWP results and giving opinions based on their experience. It is obvious that using a suitable post-processing algorithm for simulating weather consultation is an interesting and important topic. MOML is a post-processing method based on machine learning, which matches NWP forecasts against observations through a regression function. By adopting different feature engineering of datasets and training periods, the observational and model data can be processed into the corresponding training set and test set. The MOML regression function uses an existing machine learning algorithm with the processed dataset to revise the output of NWP models combined with the observations, so as to improve the results of weather forecasts. To test the new approach for grid temperature forecasts, the 2-m surface air temperature in the Beijing area from the ECMWF model is used. MOML with different feature engineering is compared against the ECMWF model and modified model output statistics (MOS) method. MOML shows a better numerical performance than the ECMWF model and MOS, especially for winter. The results of MOML with a linear algorithm, running training period, and dataset using spatial interpolation ideas, are better than others when the forecast time is within a few days. The results of MOML with the Random Forest algorithm, year-round training period, and dataset containing surrounding gridpoint information, are better when the forecast time is longer.展开更多
This paper presents a study on the statistical forecasts of typhoon tracks. Numerical models have their own systematic errors, like a bias. In order to improve the accuracy of track forecasting, a statistical model ca...This paper presents a study on the statistical forecasts of typhoon tracks. Numerical models have their own systematic errors, like a bias. In order to improve the accuracy of track forecasting, a statistical model called DLM (dynamic linear model) is applied to remove the systematic error. In the analysis of typhoons occurring over the western North Pacific in 1997 and 2000, DLM is useful as an adaptive model for the prediction of typhoon tracks.展开更多
Two intense quasi-linear mesoscale convective systems（QLMCSs） in northern China were simulated using the WRF（Weather Research and Forecasting） model and the 3D-Var（three-dimensional variational） analysis system ...Two intense quasi-linear mesoscale convective systems（QLMCSs） in northern China were simulated using the WRF（Weather Research and Forecasting） model and the 3D-Var（three-dimensional variational） analysis system of the ARPS（Advanced Regional Prediction System） model.A new method in which the lightning density is calculated using both the precipitation and non-precipitation ice mass was developed to reveal the relationship between the lightning activities and QLMCS structures.Results indicate that,compared with calculating the results using two previous methods,the lightning density calculated using the new method presented in this study is in better accordance with observations.Based on the calculated lightning densities using the new method,it was found that most lightning activity was initiated on the right side and at the front of the QLMCSs,where the surface wind field converged intensely.The CAPE was much stronger ahead of the southeastward progressing QLMCS than to the back it,and their lightning events mainly occurred in regions with a large gradient of CAPE.Comparisons between lightning and non-lightning regions indicated that lightning regions featured more intense ascending motion than non-lightning regions;the vertical ranges of maximum reflectivity between lightning and non-lightning regions were very different;and the ice mixing ratio featured no significant differences between the lightning and non-lightning regions.展开更多
Consumption of the electric power highly depends on the Season under consideration. The various means of power generation methods using renewable resources such as sunlight, wind, rain, tides, and waves are season dep...Consumption of the electric power highly depends on the Season under consideration. The various means of power generation methods using renewable resources such as sunlight, wind, rain, tides, and waves are season dependent. This paves the way for analyzing the demand for electric power based on various Seasons. Many traditional methods are utilized previously for the seasonal based electricity demand forecasting. With the development of the advanced tools, these methods are replaced by efficient forecasting techniques. In this paper, a WEKA time series forecasting is being done for the electric power demand for the three seasons such as summer, winter and rainy seasons. The monthly electric consumption data of domestic category is collected from Tamil Nadu Electricity Board (TNEB). Data collected has been pruned based on the three seasons. The WEKA learning algorithms such as Multilayer Perceptron, Support Vector Machine, Linear Regression, and Gaussian Process are used for implementation. The Mean Absolute Error (MAE) and Direction Accuracy (DA) are calculated for the WEKA learning algorithms and they are compared to find the best learning algorithm. The Support Vector Machine algorithm exhibits low Mean Absolute Error and high Direction Accuracy than other WEKA learning algorithms. Hence, the Support Vector Machine learning algorithm is proven to be the WEKA learning algorithm for seasonal based electricity demand forecasting. The need of the hour is to predict and act in the deficit power. This paper is a prelude for such activity and an eye opener in this field.展开更多
简单线性模型的时间序列预测质量通常超过Transformer等深度模型;而在具有大量通道的数据集上,深度模型尤其是多层感知器(MLP)的性能反而可超过简单线性模型。针对简单线性模型和MLP在时间序列预测中的误差功率谱差异,提出一种基于MLP...简单线性模型的时间序列预测质量通常超过Transformer等深度模型;而在具有大量通道的数据集上,深度模型尤其是多层感知器(MLP)的性能反而可超过简单线性模型。针对简单线性模型和MLP在时间序列预测中的误差功率谱差异,提出一种基于MLP的高频增强型时间序列预测模型HiFNet(High-Frequency Network)。首先,利用MLP在低频段的拟合能力;其次,通过自适应序列分解(ASD)模块及分组线性层解决MLP高频段易过拟合以及通道独立策略不能有效应对通道冗余的问题,从而增强MLP在高频段的鲁棒性;最后,对HiFNet在气象、电力和交通等领域的标准数据集上进行实验。结果表明:HiFNet的均方误差(MSE)在最佳情况下相较于NLinear、RLinear、SegRNN(Segment Recurrent Neural Network)和PatchTST(Patch Time Series Transformer)分别降低了23.6%、10.0%、35.1%和6.5%,而分组线性层通过学习通道相关性的低秩表达减轻了通道冗余的影响。展开更多
Equal Salt Deposit Density (ESDD) is a main factor to classify contamination severity and draw pollution distribution map. The precise ESDD forecasting plays an important role in the safety, economy and reliability of...Equal Salt Deposit Density (ESDD) is a main factor to classify contamination severity and draw pollution distribution map. The precise ESDD forecasting plays an important role in the safety, economy and reliability of power system. To cope with the problems existing in the ESDD predicting by multivariate linear regression (MLR), back propagation (BP) neural network and least squares support vector machines (LSSVM), a nonlinear combination forecasting model based on wavelet neural network (WNN) for ESDD is proposed. The model is a WNN with three layers, whose input layer has three neurons and output layer has one neuron, namely, regarding the ESDD forecasting results of MLR, BP and LSSVM as the inputs of the model and the observed value as the output. In the interest of better reflection of the influence of each single forecasting model on ESDD and increase of the accuracy of ESDD prediction, Morlet wavelet is used to con-struct WNN, error backpropagation algorithm is adopted to train the network and genetic algorithm is used to determine the initials of the parameters. Simulation results show that the accuracy of the proposed combina-tion ESDD forecasting model is higher than that of any single model and that of traditional linear combina-tion forecasting (LCF) model. The model provides a new feasible way to increase the accuracy of pollution distribution map of power network.展开更多
文摘This paper presents an analysis to forecast the loads of an isolated area where the history of load is not available or the history may not represent the realistic demand of electricity. The analysis is done through linear regression and based on the identification of factors on which electrical load growth depends. To determine the identification factors, areas are selected whose histories of load growth rate known and the load growth deciding factors are similar to those of the isolated area. The proposed analysis is applied to an isolated area of Bangladesh, called Swandip where a past history of electrical load demand is not available and also there is no possibility of connecting the area with the main land grid system.
文摘The fuzzy linear regression forecasting model is deduced from the symmetric triangular fuzzy number. With the help of the degree of fitting and the measure of fuzziness, the determination of symmetric triangular fuzzy numbers is changed into a problem of solving linear programming.
文摘The present paper discusses the application of localized linear models for the prediction of hourly PM10 concentration values. The advantages of the proposed approach lies in the clustering of the data based on a common property and the utilization of the target variable during this process, which enables the development of more coherent models. Two alternative localized linear modelling approaches are developed and compared against benchmark models, one in which data are clustered based on their spatial proximity on the embedding space and one novel approach in which grouped data are described by the same linear model. Since the target variable is unknown during the prediction stage, a complimentary pattern recognition approach is developed to account for this lack of information. The application of the developed approach on several PM10 data sets from the Greater Athens Area, Helsinki and London monitoring networks returned a significant reduction of the prediction error under all examined metrics against conventional forecasting schemes such as the linear regression and the neural networks.
文摘时间序列预测在能源管理、交通流量和气象分析等多个实际场景中具有重要应用价值。然而,时间序列数据中存在的分布漂移(Distribution Shift)与长程依赖(Long-term Dependency)仍限制了传统方法与现有深度学习模型在长期预测中的表现。为此,提出了一种名为D-LINet(Dual-Normalization and Linear Integration Network)的创新模型。该模型结合了Dish-TS(Distribution Shift in Time Series Forecasting)框架的分布归一化能力与线性映射的高效性,并采用双向归一化与双线性层的设计,有效缓解输入与输出空间的分布偏移,增强了对周期性与趋势性特征的捕捉能力。在多个真实数据集上对D-LINet的预测性能进行了全面评估。结果显示,在短期与长期预测中,D-LINet的均方误差和平均绝对误差均显著优于主流模型(如Transformer,Informer,Autoformer和DLinear)。此外,实验还探讨了输入窗口长度及先验知识的引入对预测性能的影响,为后续模型优化提供了重要指导。该研究针对复杂分布漂移问题提出了新的解决思路,并有助于提升时间序列预测的精度与稳健性。
文摘The 3-hour-interval prediction of ground-level temperature from +00 h out to +45 h in South Korea (38 stations) is performed using the DLM (dynamic linear model) in order to eliminate the systematic error of numerical model forecasts. Numerical model forecasts and observations are used as input values of the DLM. According to the comparison of the DLM forecasts to the KFM (Kalman filter model) forecasts with RMSE and bias, the DLM is useful to improve the accuracy of prediction.
文摘One of the principal issues related to hydrologic models for prediction of runoff is the estimation of extreme values (floods). It is well understood that unless the models capture the dynamics of rainfall-runoff process, the improvement in prediction of such extremes is far from reality. In this paper, it is proposed to develop a dual (combined and paralleled) artificial neural network (D-ANN), which aims to improve the models performance, especially in terms of extreme values. The performance of the proposed dual-ANN model is compared with that of feed forward ANN (FF-ANN) model, the later being the most common ANN model used in hydrologic literature. The forecasting exercise is carried out for hourly river flow data of Kolar Basin, India. The results of the comparison indicate that the D-ANN model performs better than the FF-ANN model.
基金supported by the National Key Research and Development Program of China (Grant Nos. 2018YFF0300104 and 2017YFC0209804)the National Natural Science Foundation of China (Grant No. 11421101)Beijing Academy of Artifical Intelligence (BAAI)
文摘In this paper, the model output machine learning (MOML) method is proposed for simulating weather consultation, which can improve the forecast results of numerical weather prediction (NWP). During weather consultation, the forecasters obtain the final results by combining the observations with the NWP results and giving opinions based on their experience. It is obvious that using a suitable post-processing algorithm for simulating weather consultation is an interesting and important topic. MOML is a post-processing method based on machine learning, which matches NWP forecasts against observations through a regression function. By adopting different feature engineering of datasets and training periods, the observational and model data can be processed into the corresponding training set and test set. The MOML regression function uses an existing machine learning algorithm with the processed dataset to revise the output of NWP models combined with the observations, so as to improve the results of weather forecasts. To test the new approach for grid temperature forecasts, the 2-m surface air temperature in the Beijing area from the ECMWF model is used. MOML with different feature engineering is compared against the ECMWF model and modified model output statistics (MOS) method. MOML shows a better numerical performance than the ECMWF model and MOS, especially for winter. The results of MOML with a linear algorithm, running training period, and dataset using spatial interpolation ideas, are better than others when the forecast time is within a few days. The results of MOML with the Random Forest algorithm, year-round training period, and dataset containing surrounding gridpoint information, are better when the forecast time is longer.
基金the project"A study on improving forecast skill using a su-percomputer"of Meteorological Research Institute,KMA,2001.
文摘This paper presents a study on the statistical forecasts of typhoon tracks. Numerical models have their own systematic errors, like a bias. In order to improve the accuracy of track forecasting, a statistical model called DLM (dynamic linear model) is applied to remove the systematic error. In the analysis of typhoons occurring over the western North Pacific in 1997 and 2000, DLM is useful as an adaptive model for the prediction of typhoon tracks.
基金supported jointly by the National Key Basic Research and Development (973) Program of China (Grant No. 2014CB441401)the National Natural Science Foundation of China (Grant Nos. 41405007, 41175043, 41475002, and 41205027)
文摘Two intense quasi-linear mesoscale convective systems（QLMCSs） in northern China were simulated using the WRF（Weather Research and Forecasting） model and the 3D-Var（three-dimensional variational） analysis system of the ARPS（Advanced Regional Prediction System） model.A new method in which the lightning density is calculated using both the precipitation and non-precipitation ice mass was developed to reveal the relationship between the lightning activities and QLMCS structures.Results indicate that,compared with calculating the results using two previous methods,the lightning density calculated using the new method presented in this study is in better accordance with observations.Based on the calculated lightning densities using the new method,it was found that most lightning activity was initiated on the right side and at the front of the QLMCSs,where the surface wind field converged intensely.The CAPE was much stronger ahead of the southeastward progressing QLMCS than to the back it,and their lightning events mainly occurred in regions with a large gradient of CAPE.Comparisons between lightning and non-lightning regions indicated that lightning regions featured more intense ascending motion than non-lightning regions;the vertical ranges of maximum reflectivity between lightning and non-lightning regions were very different;and the ice mixing ratio featured no significant differences between the lightning and non-lightning regions.
文摘Consumption of the electric power highly depends on the Season under consideration. The various means of power generation methods using renewable resources such as sunlight, wind, rain, tides, and waves are season dependent. This paves the way for analyzing the demand for electric power based on various Seasons. Many traditional methods are utilized previously for the seasonal based electricity demand forecasting. With the development of the advanced tools, these methods are replaced by efficient forecasting techniques. In this paper, a WEKA time series forecasting is being done for the electric power demand for the three seasons such as summer, winter and rainy seasons. The monthly electric consumption data of domestic category is collected from Tamil Nadu Electricity Board (TNEB). Data collected has been pruned based on the three seasons. The WEKA learning algorithms such as Multilayer Perceptron, Support Vector Machine, Linear Regression, and Gaussian Process are used for implementation. The Mean Absolute Error (MAE) and Direction Accuracy (DA) are calculated for the WEKA learning algorithms and they are compared to find the best learning algorithm. The Support Vector Machine algorithm exhibits low Mean Absolute Error and high Direction Accuracy than other WEKA learning algorithms. Hence, the Support Vector Machine learning algorithm is proven to be the WEKA learning algorithm for seasonal based electricity demand forecasting. The need of the hour is to predict and act in the deficit power. This paper is a prelude for such activity and an eye opener in this field.
文摘简单线性模型的时间序列预测质量通常超过Transformer等深度模型;而在具有大量通道的数据集上,深度模型尤其是多层感知器(MLP)的性能反而可超过简单线性模型。针对简单线性模型和MLP在时间序列预测中的误差功率谱差异,提出一种基于MLP的高频增强型时间序列预测模型HiFNet(High-Frequency Network)。首先,利用MLP在低频段的拟合能力;其次,通过自适应序列分解(ASD)模块及分组线性层解决MLP高频段易过拟合以及通道独立策略不能有效应对通道冗余的问题,从而增强MLP在高频段的鲁棒性;最后,对HiFNet在气象、电力和交通等领域的标准数据集上进行实验。结果表明:HiFNet的均方误差(MSE)在最佳情况下相较于NLinear、RLinear、SegRNN(Segment Recurrent Neural Network)和PatchTST(Patch Time Series Transformer)分别降低了23.6%、10.0%、35.1%和6.5%,而分组线性层通过学习通道相关性的低秩表达减轻了通道冗余的影响。
文摘Equal Salt Deposit Density (ESDD) is a main factor to classify contamination severity and draw pollution distribution map. The precise ESDD forecasting plays an important role in the safety, economy and reliability of power system. To cope with the problems existing in the ESDD predicting by multivariate linear regression (MLR), back propagation (BP) neural network and least squares support vector machines (LSSVM), a nonlinear combination forecasting model based on wavelet neural network (WNN) for ESDD is proposed. The model is a WNN with three layers, whose input layer has three neurons and output layer has one neuron, namely, regarding the ESDD forecasting results of MLR, BP and LSSVM as the inputs of the model and the observed value as the output. In the interest of better reflection of the influence of each single forecasting model on ESDD and increase of the accuracy of ESDD prediction, Morlet wavelet is used to con-struct WNN, error backpropagation algorithm is adopted to train the network and genetic algorithm is used to determine the initials of the parameters. Simulation results show that the accuracy of the proposed combina-tion ESDD forecasting model is higher than that of any single model and that of traditional linear combina-tion forecasting (LCF) model. The model provides a new feasible way to increase the accuracy of pollution distribution map of power network.
