Designers are required to plan for future expansion and also to estimate the grid's future utilization. This means that an effective modeling and forecasting technique, which will use efficiently the information c...Designers are required to plan for future expansion and also to estimate the grid's future utilization. This means that an effective modeling and forecasting technique, which will use efficiently the information contained in the available data, is required, so that important data properties can be extracted and projected into the future. This study proposes an adaptive method based on the multi-model partitioning algorithm (MMPA), for short-term electricity load forecasting using real data. The grid's utilization is initially modeled using a multiplicative seasonal ARIMA (autoregressive integrated moving average) model. The proposed method uses past data to learn and model the normal periodic behavior of the electric grid. Either ARMA (autoregressive moving average) or state-space models can be used for the load pattern modeling. Load anomalies such as unexpected peaks that may appear during the summer or unexpected faults (blackouts) are also modeled. If the load pattern does not match the normal be-havior of the load, an anomaly is detected and, furthermore, when the pattern matches a known case of anomaly, the type of anomaly is identified. Real data were used and real cases were tested based on the measurement loads of the Hellenic Public Power Cooperation S.A., Athens, Greece. The applied adaptive multi-model filtering algorithm identifies successfully both normal periodic behavior and any unusual activity of the electric grid. The performance of the proposed method is also compared to that produced by the ARIMA model.展开更多
Considering that the prediction accuracy of the traditional traffic flow forecasting model is low,based on kernel adaptive filter(KAF)algorithm,kernel least mean square(KLMS)algorithm and fixed-budget kernel recursive...Considering that the prediction accuracy of the traditional traffic flow forecasting model is low,based on kernel adaptive filter(KAF)algorithm,kernel least mean square(KLMS)algorithm and fixed-budget kernel recursive least-square(FB-KRLS)algorithm are presented for online adaptive prediction.The computational complexity of the KLMS algorithm is low and does not require additional solution paradigm constraints,but its regularization process can solve the problem of regularization performance degradation in high-dimensional data processing.To reduce the computational complexity,the sparse criterion is introduced into the KLMS algorithm.To further improve forecasting accuracy,FB-KRLS algorithm is proposed.It is an online learning method with fixed memory budget,and it is capable of recursively learning a nonlinear mapping and changing over time.In contrast to a previous approximate linear dependence(ALD)based technique,the purpose of the presented algorithm is not to prune the oldest data point in every time instant but it aims to prune the least significant data point,thus suppressing the growth of kernel matrix.In order to verify the validity of the proposed methods,they are applied to one-step and multi-step predictions of traffic flow in Beijing.Under the same conditions,they are compared with online adaptive ALD-KRLS method and other kernel learning methods.Experimental results show that the proposed KAF algorithms can improve the prediction accuracy,and its online learning ability meets the actual requirements of traffic flow and contributes to real-time online forecasting of traffic flow.展开更多
This paper introduces the preconditioned methods for Space-Time Adaptive Processing(STAP).Using the Block-Toeplitz-Toeplitz-Block(BTTB)structure of the clutter-plus-noise covari-ance matrix,a Block-Circulant-Circulant...This paper introduces the preconditioned methods for Space-Time Adaptive Processing(STAP).Using the Block-Toeplitz-Toeplitz-Block(BTTB)structure of the clutter-plus-noise covari-ance matrix,a Block-Circulant-Circulant-Block(BCCB)preconditioner is constructed.Based on thepreconditioner,a Preconditioned Multistage Wiener Filter(PMWF)which can be implemented by thePreconditioned Conjugate Gradient(PCG)method is proposed.Simulation results show that thePMWF has faster convergence rate and lower processing rank compared with the MWF.展开更多
This paper adopts the concept of dynamic feedback systems to model the behavior of financial markets, or more specifically, the stock market from a dynamic system point of view. Based on a feedback adaptation scheme, ...This paper adopts the concept of dynamic feedback systems to model the behavior of financial markets, or more specifically, the stock market from a dynamic system point of view. Based on a feedback adaptation scheme, the authors model the movement of a stock market index within a framework that is composed of an internal dynamic model and an adaptive filter. The output-error model is adopted as the internal model whereas the adaptive filter is a time-varying state space model with instrumental variables. Its input-output behavior, and internal as well as external forces are then identified. Special attention has also been paid to the recent financial crisis by examining the movement of Dow Jones Industrial Average （DJIA） as an example to illustrate the advantage of the proposed framework. Supported by time-varying causality tests, five influential factors from economic and sentiment aspects are introduced as the input of this framework. Testing results show that the proposed framework has a much better prediction performance than the existing methods, especially in complicated economic situations. An application of this framework is also presented with focuses on forecasting the turning periods of the market trend. Realizing that a market trend is about to change when the external force begins to exhibit clear patterns in its frequency responses, the authors develop a set of rules to recognize this kind of clear patterns. These rules work well for stock indexes from US, China and Singapore.展开更多
文摘Designers are required to plan for future expansion and also to estimate the grid's future utilization. This means that an effective modeling and forecasting technique, which will use efficiently the information contained in the available data, is required, so that important data properties can be extracted and projected into the future. This study proposes an adaptive method based on the multi-model partitioning algorithm (MMPA), for short-term electricity load forecasting using real data. The grid's utilization is initially modeled using a multiplicative seasonal ARIMA (autoregressive integrated moving average) model. The proposed method uses past data to learn and model the normal periodic behavior of the electric grid. Either ARMA (autoregressive moving average) or state-space models can be used for the load pattern modeling. Load anomalies such as unexpected peaks that may appear during the summer or unexpected faults (blackouts) are also modeled. If the load pattern does not match the normal be-havior of the load, an anomaly is detected and, furthermore, when the pattern matches a known case of anomaly, the type of anomaly is identified. Real data were used and real cases were tested based on the measurement loads of the Hellenic Public Power Cooperation S.A., Athens, Greece. The applied adaptive multi-model filtering algorithm identifies successfully both normal periodic behavior and any unusual activity of the electric grid. The performance of the proposed method is also compared to that produced by the ARIMA model.
基金National Natural Science Foundation of China(No.51467008)
文摘Considering that the prediction accuracy of the traditional traffic flow forecasting model is low,based on kernel adaptive filter(KAF)algorithm,kernel least mean square(KLMS)algorithm and fixed-budget kernel recursive least-square(FB-KRLS)algorithm are presented for online adaptive prediction.The computational complexity of the KLMS algorithm is low and does not require additional solution paradigm constraints,but its regularization process can solve the problem of regularization performance degradation in high-dimensional data processing.To reduce the computational complexity,the sparse criterion is introduced into the KLMS algorithm.To further improve forecasting accuracy,FB-KRLS algorithm is proposed.It is an online learning method with fixed memory budget,and it is capable of recursively learning a nonlinear mapping and changing over time.In contrast to a previous approximate linear dependence(ALD)based technique,the purpose of the presented algorithm is not to prune the oldest data point in every time instant but it aims to prune the least significant data point,thus suppressing the growth of kernel matrix.In order to verify the validity of the proposed methods,they are applied to one-step and multi-step predictions of traffic flow in Beijing.Under the same conditions,they are compared with online adaptive ALD-KRLS method and other kernel learning methods.Experimental results show that the proposed KAF algorithms can improve the prediction accuracy,and its online learning ability meets the actual requirements of traffic flow and contributes to real-time online forecasting of traffic flow.
基金the Innovation Foundation of NUDT forPh.D.graduates.
文摘This paper introduces the preconditioned methods for Space-Time Adaptive Processing(STAP).Using the Block-Toeplitz-Toeplitz-Block(BTTB)structure of the clutter-plus-noise covari-ance matrix,a Block-Circulant-Circulant-Block(BCCB)preconditioner is constructed.Based on thepreconditioner,a Preconditioned Multistage Wiener Filter(PMWF)which can be implemented by thePreconditioned Conjugate Gradient(PCG)method is proposed.Simulation results show that thePMWF has faster convergence rate and lower processing rank compared with the MWF.
文摘This paper adopts the concept of dynamic feedback systems to model the behavior of financial markets, or more specifically, the stock market from a dynamic system point of view. Based on a feedback adaptation scheme, the authors model the movement of a stock market index within a framework that is composed of an internal dynamic model and an adaptive filter. The output-error model is adopted as the internal model whereas the adaptive filter is a time-varying state space model with instrumental variables. Its input-output behavior, and internal as well as external forces are then identified. Special attention has also been paid to the recent financial crisis by examining the movement of Dow Jones Industrial Average （DJIA） as an example to illustrate the advantage of the proposed framework. Supported by time-varying causality tests, five influential factors from economic and sentiment aspects are introduced as the input of this framework. Testing results show that the proposed framework has a much better prediction performance than the existing methods, especially in complicated economic situations. An application of this framework is also presented with focuses on forecasting the turning periods of the market trend. Realizing that a market trend is about to change when the external force begins to exhibit clear patterns in its frequency responses, the authors develop a set of rules to recognize this kind of clear patterns. These rules work well for stock indexes from US, China and Singapore.
