27 February 2019 in IEEE Access
The volatility and uncertainty of wind power often affect the quality of electric energy, the security of the power grid, the stability of the power system and the fluctuation of the power market. In this case, the research on wind power forecasting is of great significance for ensuring the better development of wind power grids and the higher quality of electric energy. Therefore, a lot of new forecasting methods have been put forward. In this paper, a new forecasting model based on convolution neural network and LightGBM is constructed. The procedure is shown as follows, firstly, we construct new feature sets by analyzing the characteristics of the raw data on the time series from the wind field and adjacent wind field. Secondly, the convolutional neural network (CNN) is proposed to extract information from input data, and the network parameters are adjusted by comparing the actual results. After that, in consideration of the limitations of the single-convolution model in predicting wind power, we innovatively integrated LightGBM classification algorithm at the model to improve the forecasting accuracy and robustness. Finally, compared with the existing support vector machines, LightGBM and CNN, the fusion model has better performance in accuracy and efficiency.
9 December 2020 in IEEE Transactions on Instrumentation and Measurement
As a fast and non-intrusive measurement and visualization technique, Electrical Capacitance Tomography (ECT) is rapidly expanding its applications in the research on multiphase flow, fluidization, drying, combustion, and so on. However, the marked unevenness of the sensitivity maps sometimes causes unexpected effects in imaging reconstruction, particularly in three-dimension cases. To exploit the positive potential of this phenomenon, the authors proposed an image fusion method using the data from two units of ECT sensors in this study. This method is used in image fusion on the reconstructed images for a planar sensor and a cylindrical sensor. In contrast to the conventional fusion models that use fixed weight factors for two sources of data, our model forges weight functions that are set preference the strength of the sensitivity maps. The new algorithm is implemented by first extracting the characteristic information out of the ECT images using the Latent Low-Rank Representation and then performing a fusion algorithm with linear weight functions in preference to the significance of the sensitivity maps. The simulation results show that the algorithm effectively retains the advantages of the two units of sensors and mutually compensates the weak points of theirs, and significantly improves the reconstruction quality. The fusion image quality by the new method can response the real situation better in different heights. The results imply that the data fusion might to a significant extend amend the weakness of ECT caused by the uneven sensitivity maps.