Search

The implementation of rainwater harvesting (RWH) has emerged as a key strategy to cope with the water crisis in urban areas. However, the required design parameters of the RWH system in different buildings vary widely because of the differences in architectural characteristics. In this paper, the relationships between the building characteristics parameters and the optimal tank (Tc) capacity are investigated by simulating the RWH system for more than 120 buildings at different locations in Guangzhou. Explicit expressions relating the optimal storage volume of RWH systems to building characteristics are derived based on nonlinear regression analysis. A set of bivariate exponential equations for estimating the optimal tank size under different building conditions was obtained. The model has a MER of less than 10% for buildings with a C/D ratio between 17.5 and 85. In Guangzhou, the unit cost of an RWH system can be reduced to CNY 2.62/m2 with optimally designed tanks. Highlight: A generalization equation is proposed to assist in the design of cisterns for RWH systems.The validity range of the equation is verified.The derived analytical equation is adequate to determine the cistern volume for medium‐sized buildings. [ABSTRACT FROM AUTHOR]

Rainwater harvesting (RWH) systems are one of the most promising technologies for water supply, but the economic viability often barricades their implementation. To assess the economic viability of decentralized RWH systems at a regional level, this paper develops a generic method to investigate the effect of variations in building characteristics on the economic performance of rainwater harvesting systems in regions with fixed water tariffs and rainfall distribution characteristics. We simulated the financial efficiency (expressed as a benefit-cost ratio) of a large number of decentralized RWH systems in Guangzhou, China. We found that the financial efficiency of the RWH systems is closely related to the catchment fraction (i.e., the ratio of rainfall catchment surface to total floor area). Based on this dimensionless parameter, we derived explicit equations expressing the financial efficiency of RWH systems using a nonlinear regression method. In addition, we validated the analytical equations by the root mean square error test, normality test, and error distribution test. This analytical solution provides a simple and generic method for forecasting the investment potential of rainwater harvesting systems. This methodology can also be adapted to other regions where the variations in local water price and rainfall data need to be considered. [ABSTRACT FROM AUTHOR]

The spatial decomposition of demographic data at a fine resolution is a classic and crucial problem in the field of geographical information science. The main objective of this study was to compare twelve well-known machine learning regression algorithms for the spatial decomposition of demographic data with multisource geospatial data. Grid search and cross-validation methods were used to ensure that the optimal model parameters were obtained. The results showed that all the global regression algorithms used in the study exhibited acceptable results, besides the ordinary least squares (OLS) algorithm. In addition, the regularization method and the subsetting method were both useful for alleviating overfitting in the OLS model, and the former was better than the latter. The more competitive performance of the nonlinear regression algorithms than the linear regression algorithms implies that the relationship between population density and influence factors is likely to be non-linear. Among the global regression algorithms used in the study, the best results were achieved by the k-nearest neighbors (KNN) regression algorithm. In addition, it was found that multi-sources geospatial data can improve the accuracy of spatial decomposition results significantly, and thus the proposed method in our study can be applied to the study of spatial decomposition in other areas. [ABSTRACT FROM AUTHOR]

Cooling load prediction is a critically important technology for ensuring energy conservation in cooling systems. Examples include advance air conditioning control, chiller startup, and controlling the number of chilled water pumps in operation. Many existing load prediction methods are normally arranged into three categories: energy simulation, artificial intelligence, and regression analysis. In contrast to the first two categories, regression analysis is practical and easy to implement in real applications. However, cooling load data always exhibits nonlinear, dynamic features, thus it is very difficult to accurately predict cooling loads with linear regression models. Thus, a multiple nonlinear regression (MNR) model for cooling systems in public buildings was used to accurately predict short term cooling loads in this study. The key variables in the MNR model are selected based on sensitivity analysis, and calibration methods can be used to enhance the prediction accuracy. The prediction process can be divided into three steps. The first step requires determining the weight coefficients on the input variables based on historical load and weather data. Second, the initially predicted load is obtained using the weight coefficients and predicted input variables. Finally, the initially predicted results were calibrated using relative calibration methods. The initial load prediction and final calibration are based on data from the same day, which is taken as a reference. The performance of the established MNR model is validated against a real case in Guangzhou using measured cooling load data and weather data. A case study shows that the accuracy of the final calibrated load is higher than the initially predicted load. The MNR model also produces more accurate results compared with three traditional regression models. The accurate cooling load prediction method can be used as an operation control strategy in real applications. [ABSTRACT FROM AUTHOR]

• The energy consumption of three lines and a station of Guangzhou Metro is revealed. • The prediction of electricity consumption of urban rail transit is conducted. • The performances of the BNFR and SVR models are compared. • The SVR model is superior to the BNFR model in terms of accuracy. Predicting the energy consumption of urban rail transit is conducive to reducing energy consumption in the subway system. Therefore, binary nonlinear fitting regression (BNFR) and support vector regression (SVR) models are developed to predict total electricity, traction electricity, and heating ventilation air conditioning (HVAC) system electricity consumption in subway lines as well as the electricity consumption of chillers in a subway station. The two models are compared in terms of accuracy, and the results demonstrate that the SVR model is superior to the BNFR model. The prediction accuracies of traction electricity and total electricity consumption in subway lines are high. By contrast, the prediction accuracy of the HVAC system electricity consumption in subway lines is low. This is due to numerous factors aside from outdoor temperature, operation mileages, and passenger flow, which can influence the HVAC system electricity. Thus, the influencing factors should further be investigated to increase the prediction accuracy. The electricity consumption of chillers in subway station can be predicted with the comprehensive consideration of indoor and outdoor temperatures and humidity levels, passenger flows, timetable of trains, power of new draught fans, and chilling of strong electricity rooms (SERs). [ABSTRACT FROM AUTHOR]

• Five prediction models (MLR, AR, ARX, MNR, and BP models) were developed for short term cooling load prediction. • The prediction accuracy on different training time scales (T-1 to T-6) were analyzed, and was evaluated by five evaluation indices. • Performance results showed that prediction accuracy of MNR was highest than MLR, AR, ARX and BP models when using short time training scale T-1. • For short time scale prediction, MNR model requires less training data, low hardware requirements and low computational complexity, which can be easily used to quickly on-site load prediction. Data-mining approaches for improving building cooling load predictions are presented and analyzed on different training time scales (T-1 to T-6) in this paper. Multiple linear regression (MLR), auto regression (AR), autoregressive exogenous (ARX) model, multiple nonlinear regression (MNR), and a back propagation neural network (BP). In this study, a real case in Guangzhou was examined to verify the performance of the prediction models on different time scales using measured cooling load and weather data. Based on the sensitivity analysis, the variable with largest sensitivity is selected as the input to the proposed MNR model, which can ensure prediction accuracy while increasing prediction speed. The case study shows that the prediction from the MNR model is comparable to or better than that from the BP model on different training time scales. The results also show that the MNR model could provide satisfactory prediction accuracy with less data (e.g., 3 d of weather and historical load data), while the BP model requires at least one month of training data to provide the same prediction accuracy. An analysis of the prediction performance results shows that the prediction accuracy from the regression models will decrease and the prediction accuracy from the BP model gradually increases as the training time scale increases. Surprisingly, the prediction accuracy with the MNR model is superior to that from the BP model with T-1 to T-4 training time scales. Finally, it is shown that the MNR model with the limited variable as inputs provides higher prediction accuracy than four other models with training time scales T-1 to T-4. The MNR model requires less training data, simpler hardware, and low computational complexity, making it preferable for quick on-site load prediction. [ABSTRACT FROM AUTHOR]