#### DMCA

## Particle Swarm Optimization-based LS-SVM for Building Cooling Load Prediction

### Citations

12873 | Statistical Learning Theory
- Vapnik
- 1998
(Show Context)
Citation Context ...ilding load using them.sRecently, a novel type of learning machine, calledssupport vector machine (SVM), has been receivingsincreasing attention. SVM was developed by Vapnik andshis coworkers in 1995 =-=[7]-=-, and it is based on the structuresrisk minimization (SRM) principle that seeks to minimizesan upper bound of the generalization error consisting ofsthe sum of the training error and a confidence inte... |

3521 | Particle swarm optimization, in
- Kennedy, Eberhart
- 1995
(Show Context)
Citation Context ...training phase.sBasic principle of PSOsPSO is a stochastic optimization technique introducedsrecently by Kennedy and Eberhart, which is inspired byssocial behavior of bird flocking and fish schooling =-=[14,s15]-=-. Similar to other evolutionary computation algorithmsssuch as genetic algorithms, PSO is a population-basedssearch method that exploits a population of individuals tosprobe promising region of the se... |

3327 |
Time Series Analysis Forecasting and Control. 2 nd edition
- Box, Jenkins
- 1976
(Show Context)
Citation Context ... T l ∈= ,),...,,( 21 , thesregularization constant γ > 0 is included to control thesbias-variance trade-off. The above statement is in fact thessame formulation as is used in case of ridge regressions=-=[21]-=- in the feature space defined by )(⋅ϕ . Note that inssome cases w becomes infinite dimension, and the abovesproblem formulation cannot be used to solve the problem.sTherefore, we perform the computati... |

783 | A new optimizer using particle swarm theory, in - Eberchart, Kennedy - 1995 |

188 |
The Particle Swarm: Social Adaptation of Knowledge
- Kennedy
- 1997
(Show Context)
Citation Context ...A) were employed to choose the parameters of a SVMsmodel, and the improved model offers a superiorsperformance to ordinary regression SVM model[11,11].sThe particle swarm optimization (PSO) algorithm =-=[13]-=-, asrelatively new evolutionary computation (EC) stochasticstechnique, can also be used as an excellent optimizerswhich originated as a simulation of the food-searchingsbehavior of birds. Similar to E... |

73 | On the computation of all global minimizers through particle swarm optimization - Parsopoulos, Vrahatis - 2004 |

58 | Particle swarm optimization method constrained optimization problems
- Parsopoulos, Vrahatis
- 2002
(Show Context)
Citation Context ...imum in presence of many local optima, simplesprogramming and adaptability with constrained problems.sPSO has showed to be promising for solving varioussengineering problems such as automatic control =-=[18]-=-,santenna design [19], and inverse problems [20]. Thesgeneral principles for the PSO algorithm are stated assfollows:sLet us consider a swarm of size n. Each particle Pi (i =s1, 2, . . . , n) from the... |

42 | Parallel Global Optimization with the Particle Swarm Algorithm
- Schutte, Reinbolt, et al.
- 2004
(Show Context)
Citation Context ...any local optima, simplesprogramming and adaptability with constrained problems.sPSO has showed to be promising for solving varioussengineering problems such as automatic control [18],santenna design =-=[19]-=-, and inverse problems [20]. Thesgeneral principles for the PSO algorithm are stated assfollows:sLet us consider a swarm of size n. Each particle Pi (i =s1, 2, . . . , n) from the swarm is characteriz... |

12 | Cooling load prediction for buildings using general regression neural network - Ben-Nakhi, Mahmoud |

11 | Evaluation of Support Vector Machine Based Forecasting Tool in Electricity Price Forecasting for Australian National Electricity Market Participants - Sansom, Downs, et al. - 2003 |

5 | Hourly cooling load prediction by a combined forecasting model based on analytic hierarchy process - Yao, Lian, et al. - 2004 |

5 |
A Gaussian maximum likelihood formulation for short-term forecasting of traffic flow
- Lin
- 2001
(Show Context)
Citation Context ...maximum likelihood (GML) approachsLin’s GML-based model makes use of both historicalsand real time information in an integrated way by usingstwo key variables: cooling load and cooling loadsincrement =-=[22]-=-. Let Xi, (i = 0, 1, 2, …, n) be consecutivesobservations of the building cooling load obtained at timesi. Let Yi= Xi - Xi-1 be the flow increment. Assuming thatsthese two variables are normally distr... |

4 |
Hybrid evolutionary computation for the development of pollution prevention and control strategies
- Tan
- 2007
(Show Context)
Citation Context ...ogramming and adaptability with constrained problems.sPSO has showed to be promising for solving varioussengineering problems such as automatic control [18],santenna design [19], and inverse problems =-=[20]-=-. Thesgeneral principles for the PSO algorithm are stated assfollows:sLet us consider a swarm of size n. Each particle Pi (i =s1, 2, . . . , n) from the swarm is characterized by: 1) itsscurrent posit... |

3 |
Monthly-averaged cooling load calculations-residential and small commercial buildings
- Bida, JF
- 1987
(Show Context)
Citation Context ... an approach to integratesand optimize the heating, ventilating, and airconditioning (HVAC) system cooling supply systemsefficiently, based on which the air-conditioning supplysmatches the demand well=-=[1]-=-. What is more, it is alsosuseful for HVAC operations including adjusting thesstarting time of cooling to meet start-up loads,sminimizing or limiting the electric on-peak demand,soptimizing the costs ... |

3 | Wong LT. Cooling load calculations in subtropical climate. Build Environ - KW |

3 |
Evolutionary tuning of multiple
- Friedrichs, Igel
- 2005
(Show Context)
Citation Context ...ues and needs to locate the intervalsof feasible solution and a suitable sampling step. Becausesof the computational complexity, grid search is onlyssuitable for the adjustment of very few parameters =-=[10]-=-.sIn farther researches, some intelligent algorithms suchsas evolution algorithms (EA) and genetic algorithmss(GA) were employed to choose the parameters of a SVMsmodel, and the improved model offers ... |

3 |
Short-term traffic flow prediction based on ratio-median lengths of intervals two-factors high-order fuzzy time series
- Wang
- 2007
(Show Context)
Citation Context ...l results ofsSVR predictions encourage our research in using SVR forscooling load forecasting modeling. SVM possess greatspotential and superior performance as is appeared insmany previous researches =-=[23]-=-, the results guaranteesglobal minima. SVM can solve some flaws of the neuralsnetworks, and has many unique advantages in the fieldssof small samples and high-dimensional nonlinearsmanifested.sThere i... |

2 | A procedure for calculating cooling load due to solar radiation: the shading effects from adjacent or nearby buildings. Energy Build - Ok |

2 | Al-Johani Khalid M. Utilizing transfer function method for hourly cooling load calculations. Energy Convers Manage - MA |

2 |
Support vector machines with simulated annealing algorithms in electricity load forecasting, Energy Conversion and Management
- Ping, Wei
(Show Context)
Citation Context ...chsas evolution algorithms (EA) and genetic algorithmss(GA) were employed to choose the parameters of a SVMsmodel, and the improved model offers a superiorsperformance to ordinary regression SVM model=-=[11,11]-=-.sThe particle swarm optimization (PSO) algorithm [13], asrelatively new evolutionary computation (EC) stochasticstechnique, can also be used as an excellent optimizerswhich originated as a simulation... |

2 | Forecasting regional electricity load based on recurrent support vector machines with genetic algorithms - Ping, Wei |

1 |
Hiroshi Yoshino and Akashi Mochida, Applying support vector machine to predict hourly cooling load in the building
- Lia, Meng, et al.
(Show Context)
Citation Context ...ntages in the fieldssof small samples and high-dimensional nonlinearsmanifested.sThere is also a great deal of researches concentratingson applying SVM regression to building cooling loadsforecasting =-=[9]-=- and the forecasting accuracy outperformssother forecasting models. The selection of parameters of asSVM model is important to the accuracy of thesforecasting. However, most SVM practitioners selectst... |