Yang Shen

Continuous glucose monitoring data have strong time variability as well as complex non-stationarity and nonlinearity. The existing blood glucose concentration prediction models often overlook the impacts of residual components after multi-scale decomposition on prediction accuracy. To enhance the prediction accuracy, a new short-term glucose prediction model that integrates the double decomposition technique, nonlinear marine predator algorithm (NMPA) and deep extreme learning machine (DELM) is proposed. First of all, the initial blood glucose data are decomposed by variational mode decomposition (VMD) to reduce its complexity and non-stationarity. To make full use of the decomposed residual component, the time-varying filter empirical mode decomposition (TVF-EMD) is utilized to decompose the component, and further realize complete decomposition. Then, the NMPA algorithm is utilized to optimize the weight parameters of the DELM network to avoid any fluctuations in prediction performance, and all the decomposed subsequences are predicted separately. Finally, the output results of each model are superimposed to acquire the predicted value of blood sugar concentration. Using actual collected blood glucose concentration data for predictive analysis, the results of three patients show the following: (i) The double decomposition strategy effectively reduces the complexity and volatility of the original sequence and the residual component. Making full use of the important information implied by the residual component has the best decomposition effect; (ii) The NMPA algorithm optimizes DELM network parameters, which can effectively enhance the predictive capabilities of the network and acquire more precise predictive results; (iii) The model proposed in this paper can achieve a high prediction accuracy of 45 min in advance, and the root mean square error values are 5.2095, 4.241 and 6.3246, respectively. Compared with the other eleven models, it has the best prediction accuracy.

To meet the challenges of energy sustainability, the integrated energy system (IES) has become a key component in promoting the development of innovative energy systems. Accurate and reliable multivariate load prediction is a prerequisite for IES optimal scheduling and steady running, but the uncertainty of load fluctuation and many influencing factors increase the difficulty of forecasting. Therefore, this article puts forward a multi-energy load prediction approach of the IES, which combines the fennec fox optimization algorithm (FFA) and hybrid kernel extreme learning machine. Firstly, the comprehensive weight method is used to combine the entropy weight method and Pearson correlation coefficient, fully considering the information content and correlation, selecting the key factors affecting the prediction, and ensuring that the input features can effectively modify the prediction results. Secondly, the coupling relationship between the multi-energy load is learned and predicted using the hybrid kernel extreme learning machine. At the same time, the FFA is used for parameter optimization, which reduces the randomness of parameter setting. Finally, the approach is utilized for the measured data at Arizona State University to verify its effectiveness in multi-energy load forecasting. The results indicate that the mean absolute error (MAE) of the proposed method is 0.0959, 0.3103 and 0.0443, respectively. The root mean square error (RMSE) is 0.1378, 0.3848 and 0.0578, respectively. The weighted mean absolute percentage error (WMAPE) is only 1.915%. Compared to other models, this model has a higher accuracy, with the maximum reductions on MAE, RMSE and WMAPE of 0.3833, 0.491 and 2.8138%, respectively.