Predicting the electricity consumption of urban rail transit based on binary nonlinear fitting regression and support vector regression

Research on the non-smooth problems in the nonlinear support vector regression. A nonlinear smooth support vector regression model is proposed. Using a generalized cubic spline function approach the non-smooth part in the support vector regression model. The model of the nonlinear smooth support vector regression is solved by BFGS-Armijo. Then, the approximation accuracy and the astringency of the generalized cubic spline function to the insensitive loss function were analyzed. As a result, we found the four-order and six times spline function's approximation effect is better than other smooth functions, and the nonlinear smooth support vector regression model, which be proposed in this paper is convergent.

One of the most important characteristics contributing to the thermal management efficiency of commercial, industrial, institutional or home environments is the optimal functioning of HVAC (heating, ventilation, air conditioning) systems. In addition to using supervisor controllers for balancing comfort level in a building, the majority of today’s HVACs employ nonlinear time variance controllers when dealing with a variety of disturbances. This paper investigates both current and potential HVAC systems at Memorial University’s S. J. Carew building, St. John’s, Newfoundland. The study investigates the viability of algorithm-based supervisor fuzzy logic controllers (SFLC) for the control of the building’s four air-handling units (AHUs) used to manage the interior environment. Along with temperature, the SFLCs also control the AHUs’ fan speeds and CO2 concentrations modifying hot water and air flow rates. This work presents models of damper positions, fan speeds and globe valves that have been built in accordance with current rates of air and hot water flow in the S. J. Carew building. Based on these specifications, a novel method of SFLC adaptation using fuzzy rules has been devised. The novel system aims to better balance the performance level of the Carew building’s HVAC system on a floor-by-floor basis. The overall results indicate better overall thermal comfort levels and enhanced cost-effectiveness when using the SFLC redesign.

An accurate air-temperature prediction can provide the energy consumption and system load in advance, both of which are crucial in HVAC (heating, ventilation, air conditioning) system operation optimisation as a way of reducing energy losses, operating costs, as well as pollution and dust emissions while maintaining residents’ thermal comfort. This article presents the results of an outdoor air-temperature time-series prediction for a multifamily building with the use of artificial neural networks during the heating period (October–May). The aim of the research was to analyse in detail the created neural models with a view to select the best combination of predictors and the optimal number of neurons in a hidden layer. To meet that task, the Akaike information criterion was used. The most accurate results were obtained by MLP 3-3-1 (r = 0.986, AIC = 1300.098, SSE = 4467.109), with the ambient-air-temperature time series observed 1, 2, and 24 h before the prognostic temperature as predictors. The AIC proved to be a useful method for the optimum model selection in a machine-learning modelling. What is more, neural network models provide the most accurate prediction, when compared with LR and SVR. Additionally, the obtained temperature predictions were used in HVAC applications: entering-water temperature and indoor temperature modelling.

Short-term metro passenger flow prediction plays a great role in traffic planning and management, and it is an important prerequisite for achieving intelligent transportation. So, a novel hybrid Support Vector Regression (SVR) model based on Twice Clustering (TC) is proposed for short-term metro passenger flow prediction. The training sets and test sets are generated by TC with respect to values of passenger flow in different time periods to improve the prediction accuracy. Furthermore, each obtained cluster is decomposed by using the Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) algorithm and the Ensemble Empirical Mode Decomposition (EEMD) algorithm, respectively. The volatility of each component obtained after decomposition is further reduced. Then, the SVR model optimized by the Grey Wolf Optimization (GWO) algorithm is used to predict the decomposed components. Moreover, forecast based on one-month data from Xi’an Metro Line 2 Library Station (China). By comparing the prediction results of the TC condition, the Once Clustering (OC) condition and the non-clustering condition, it shows that the TC approach can adequately model the volatility and effectively improve the prediction accuracy. At the same time, experimental results show that the novel hybrid TC–CEEMDAN–GWO–SVR model has superior performance than Genetic Algorithm (GA) optimized SVR (SVR–GA) model and hybrid Back Propagation Neural Network (BPNN) model.

The absorption of capillary water is one of the most crucial factors in the flow of groundwater in rocks (CWA). Although meticulous experimental studies are needed to determine a rock’s CWA, predictive techniques might cut down on the expense and effort. There are various data mining methods for this purpose, but the considered algorithms in this study were not proposed so far for predicting the CWA. Different rock samples were taken for this purpose from various locations, yielding diverse rocks. For the prediction procedures, four support vector regression (SVR) models were created: a traditional SVR, two ensembled models, and a hybrid SVR model using the whale optimization technique (WOA - SVR). Results show that all models have acceptable performance in predicting the CWA with R2 larger than 0.797 and 0.806 for the training and testing data, respectively, representing the acceptable correlation between observed and predicted values. Regarding developed models, the conventional SVR model has the worst performance of all models. All statistical evaluation criteria were improved by assembling models, which present the ability of additive regression and bagging predictions in improving prediction processes. The hybrid WOA - SVR model has the best performance considering all indices. This hybrid model could also gain the lowest values of error indices between all SVR models, which leads to outperforming the WOA - SVR model compared to other methods.

Sediment phenomenon is very important in hydraulic and water resources issues. The existence of this phenomenon causes many problems in water storage. Sediment simulation in rivers helps in controlling sediment as well as reducing damages. In this study, an attempt was made to estimate the suspended sediment load using the corresponding river flow rate in the Zohreh River, Iran using the newest intelligent simulation methods. This study seeks to couple the nonlinear support vector regression (SVR) with crowd intelligence optimization algorithms. For this purpose, support vector regression was optimized using four new crowd optimization algorithms including the ant colony optimizer (ACO), the ant lion optimizer (ALO), the dragonfly algorithm (DA), and the salp swarm algorithm (SSA). Simulation was done in the two phases of train and test. Due to the integration of the nonlinear support vector regression with the optimization algorithms, the model train phase requires more time than usual situations. Therefore, in the current study, taking into account the number of different iterations including 25, 50, 100 and 200 iterations to perform the optimization of the model and tried to find the best optimizer by considering the calculated error and the run time. It was generally found that the SVR model is accurate in estimating the suspended sediment load. Finally, according to the calculated error as well as the run time, the support vector regression model optimized with the salp swarm algorithm with 25 iterations was chosen as the best model. Also, the values of R2, NSE, and RMSE for the best model in the test phase were calculated as 1, 1, and 10.2 tons per day, respectively, and the algorithm run time was 252 s.

In this study, the HVAC(Heating, Ventilation, Air Conditioning) system is designed based on the model in which the HVAC fan driving type and hybrid charging type are combined together to optimize the energy efficiency of the HVAC system by applying eco-friendly photovoltaic power generation. The efficiency increased by approximately 10% when the DC step down converter, also known as a buck converter that has a circuit with the input and output sharing the same ground, was used and applied in the HVAC blower motor. At the same time, the system was also designed to supply electric power as an auxiliary power source to the alternator and to charge the battery as a boost by making use of the LTC3814-5, which is a chip exclusively for the DC-DC step up converters. In order to confirm the efficiency of the proposed system, operating experiment and vehicle performance evaluation were conducted.

This study deals with estimation of the total and industrial sector electricity consumption based on genetic algorithm (GA) approach, and then proposes two scenarios to project future consumptions. Total electricity consumption is estimated based on gross national product (GNP), population, import and export figures of Turkey. Industrial sector electricity is calculated based on the GNP, import and export figures. Three forms of the genetic algorithm electricity demand (GAED) models for the total and two forms for the industrial electricity consumption are developed. The best-fit GAED model in terms of total minimum relative average errors between observed and estimated values is selected for future demand estimation. ‘High- and low-growth scenarios’ are proposed for predicting the future electricity consumption. Results showed that the GAED estimates the electricity demand in comparison with the other electricity demand projections. The GAED model plans electricity demand of Turkey until 2020. Copyright © 2005 John Wiley & Sons, Ltd.

This paper uses non-linear support vector regression (SVR) to model the blood volume and heart rate (HR) responses in 9 hemodynamically stable kidney failure patients during hemodialysis. Using radial bias function (RBF) kernels the non-parametric models of relative blood volume (RBV) change with time as well as percentage change in HR with respect to RBV were obtained. The e-insensitivity based loss function was used for SVR modeling. Selection of the design parameters which includes capacity (C), insensitivity region (e) and the RBF kernel parameter (sigma) was made based on a grid search approach and the selected models were cross-validated using the average mean square error (AMSE) calculated from testing data based on a k-fold cross-validation technique. Linear regression was also applied to fit the curves and the AMSE was calculated for comparison with SVR. For the model based on RBV with time, SVR gave a lower AMSE for both training (AMSE=1.5) as well as testing data (AMSE=1.4) compared to linear regression (AMSE=1.8 and 1.5). SVR also provided a better fit for HR with RBV for both training as well as testing data (AMSE=15.8 and 16.4) compared to linear regression (AMSE=25.2 and 20.1).

Nonlinear QSAR models with high-dimensional descriptor selection and SVR improve toxicity prediction and evaluation of phenols on Photobacterium phosphoreum

Image data from space-borne thermal infrared (IR) sensors are used for a variety of applications, however they are often limited by their temporal resolution (i.e., repeat coverage). To potentially increase the temporal availability of thermal image data, a study was performed to determine the extent to which thermal image data can be simulated from available atmospheric and surface data. The work conducted here explored the use of Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2) developed by The National Aeronautics and Space Administration (NASA) to predict top-of-atmosphere (TOA) thermal IR radiance globally at time scales finer than available satellite data. For this case study, TOA radiance data was derived for band 31 (10.97 μ m) of the Moderate-Resolution Imaging Spectroradiometer (MODIS) sensor. Two approaches have been followed, namely an atmospheric radiative transfer forward modeling approach and a supervised learning approach. The first approach uses forward modeling to predict TOA radiance from the available surface and atmospheric data. The second approach applied four different supervised learning algorithms to the atmospheric data. The algorithms included a linear least squares regression model, a non-linear support vector regression (SVR) model, a multi-layer perceptron (MLP), and a convolutional neural network (CNN). This research found that the multi-layer perceptron model produced the lowest overall error rates with an root mean square error (RMSE) of 1.36 W/m 2 ·sr· μ m when compared to actual Terra/MODIS band 31 image data. These studies found that for radiances above 6 W/m 2 ·sr· μ m, the forward modeling approach could predict TOA radiance to within 12 percent, and the best supervised learning approach can predict TOA to within 11 percent.

Remaining useful life estimation based on nonlinear feature reduction and support vector regression

A stacking ensemble model for predicting soil organic carbon content based on visible and near-infrared spectroscopy

Prediction of TVB-N content in eggs based on electronic nose

A method to determine the norm vacuum of duplex pressure condenser based on the support vector regression (SVR) model was proposed, The norm vacuum of condenser is usually obtained from the characteristic curve given by the manufactory. But the method is only fit for the condition that both of the flow of cooling water and exhaust steam of steam turbine are maintained at designed valves. In this paper, the factors which affected the norm vacuum of condenser were analyzed, and then a new method was studied with SVR model's parameters optimized by genetic algorithm (GA). Data from duplex pressure condenser of a 600MW unit were validated by the proposed method. The results show that the SVR model has high prediction accuracy and very strong stability. The studying results fall foundations for diagnosing the faults of low vacuum for condenser.