Weighted LS-SVM Method for Building Cooling Load Prediction

2010 ◽  
Vol 121-122 ◽  
pp. 606-612
Author(s):  
Xue Mei Li ◽  
Jia Shu Chen ◽  
Li Xing Ding
Keyword(s):  
Time Series ◽  
Time Window ◽  
Weighted Least Squares ◽  
Superior Performance ◽  
Support Vector ◽  
Model Parameters ◽  
Cooling Load ◽  
Load Prediction ◽  
On Line ◽  
Building Cooling

A number of different forecasting methods have been proposed for cooling load forecasting including historic method, real-time method, time series analysis, and artificial neural networks, but accuracy and time efficiency in prediction are a couple of contradictions to be hard to resolve for building cooling load prediction. In order to improve the prediction accuracy of cooling load time series, weighted least squares support vector machine regression (WLS-SVM) method for a chaotic cooling load prediction is proposed. In this method, a sliding time window is built and data in the sliding time window are employed to reconstruct the dynamic model. Different weights are assigned to different data in the sliding time window, and the model parameters are refreshed on-line with the rolling of the time window. The results show that the method has more superior performance than other methods like LS-SVM.

Building Cooling Load Forecasting Based on Support Vector Machines with Simulated Annealing

2010 ◽  
Vol 108-111 ◽  
pp. 1003-1008
Author(s):  
Xue Mei Li ◽  
Li Xing Ding ◽  
Jin Hu Lǔ ◽  
lan Lan Li
Keyword(s):  
Simulated Annealing ◽  
Support Vector Machines ◽  
Support Vector ◽  
Cooling Load ◽  
Load Prediction ◽  
Promising Alternative ◽  
Vector Machines ◽  
Accurate Forecast ◽  
Electricity Load ◽  
Building Cooling

Accurate forecasting of building cooling load has been one of the most important issues in the electricity industry. Recently, along with energy-saving optimal control, accurate forecast of electricity load has received increasing attention. Because of the general nonlinear mapping capabilities of forecasting, artificial neural networks have played a crucial role in forecasting electricity load. Support vector machines (SVMs) have been successfully employed to solve nonlinear regression and time series problems. In order to improve time efficiency of prediction, a new hourly cooling load prediction model and method based on Support Vector Machine in this paper. Moreover, simulated annealing (SA) algorithms were employed to choose the parameters of a SVM model. Subsequently, examples of cooling load data from Guangzhou were used to illustrate the proposed SVM-SA model. A comparison of the performance between SVM optimized by Particle Swarm Optimization (SVM-PSO) and SVM-SA is carried out. Experiments results demonstrate that SVM-SA can achieve better accuracy and generalization than the SVM-PSO. Consequently, the SVM-SA model provides a promising alternative for forecasting building load.

scholarly journals Stream Data Load Prediction for Resource Scaling Using Online Support Vector Regression

Algorithms ◽  
2019 ◽  
Vol 12 (2) ◽  
pp. 37 ◽  
Author(s):  
Zhigang Hu ◽  
Hui Kang ◽  
Meiguang Zheng
Keyword(s):  
Support Vector Regression ◽  
Real Time ◽  
Virtual Machines ◽  
Time Window ◽  
Performance Model ◽  
Streaming Data ◽  
Support Vector ◽  
Load Prediction ◽  
Stream Data ◽  
Online Support Vector Regression

A distributed data stream processing system handles real-time, changeable and sudden streaming data load. Its elastic resource allocation has become a fundamental and challenging problem with a fixed strategy that will result in waste of resources or a reduction in QoS (quality of service). Spark Streaming as an emerging system has been developed to process real time stream data analytics by using micro-batch approach. In this paper, first, we propose an improved SVR (support vector regression) based stream data load prediction scheme. Then, we design a spark-based maximum sustainable throughput of time window (MSTW) performance model to find the optimized number of virtual machines. Finally, we present a resource scaling algorithm TWRES (time window resource elasticity scaling algorithm) with MSTW constraint and streaming data load prediction. The evaluation results show that TWRES could improve resource utilization and mitigate SLA (service level agreement) violation.

scholarly journals Recurrence time distribution and temporal clustering properties of a cellular automaton modelling landslide events

2013 ◽  
Vol 20 (6) ◽  
pp. 1071-1078 ◽  
Author(s):  
E. Piegari ◽  
R. Di Maio ◽  
A. Avella
Keyword(s):  
Time Series ◽  
Probability Distribution ◽  
Cellular Automaton ◽  
Time Window ◽  
Fano Factor ◽  
Recent Analysis ◽  
Recurrence Time ◽  
Model Parameters ◽  
Reasonable Prediction ◽  
First Time

Abstract. Reasonable prediction of landslide occurrences in a given area requires the choice of an appropriate probability distribution of recurrence time intervals. Although landslides are widespread and frequent in many parts of the world, complete databases of landslide occurrences over large periods are missing and often such natural disasters are treated as processes uncorrelated in time and, therefore, Poisson distributed. In this paper, we examine the recurrence time statistics of landslide events simulated by a cellular automaton model that reproduces well the actual frequency-size statistics of landslide catalogues. The complex time series are analysed by varying both the threshold above which the time between events is recorded and the values of the key model parameters. The synthetic recurrence time probability distribution is shown to be strongly dependent on the rate at which instability is approached, providing a smooth crossover from a power-law regime to a Weibull regime. Moreover, a Fano factor analysis shows a clear indication of different degrees of correlation in landslide time series. Such a finding supports, at least in part, a recent analysis performed for the first time of an historical landslide time series over a time window of fifty years.

Sign in / Sign up

Export Citation Format

Share Document