Deep belief network based ensemble approach for cooling load forecasting of air-conditioning system

Energy ◽  
2018 ◽  
Vol 148 ◽  
pp. 269-282 ◽  
Author(s):  
Guoyin Fu
2012 ◽  
Vol 220-223 ◽  
pp. 622-625
Author(s):  
Xue Li Zhu ◽  
Bo Dong ◽  
Yong Jun Zhu

With the characteristics of non-stationarity, non-linearity, time-lag of refrigeration/ heating supplying in minds, load forecasting of central air-conditioning system is carried using time sequence analysis method. Firstly, acquisition sample data of central air-conditioning system is pretreated, and random time sequence AR model of system is formulated. Then, forecasting of AR refrigeration/heating load based on Yule-walker method is conducted. In order to enhance forecasting accuracy, crossover forecasting is introduced into the load forecasting, that is, to use vertical forecasting to follow household demands for load and horizontal forecasting to track changes of weather. Then, weight cross is made to vertical and horizontal forecasting results. Finally, refrigeration/heating load forecasting software of central air-conditioning system is developed, which is used in energy-saving monitoring and control of central air-conditioning system.


2011 ◽  
Vol 474-476 ◽  
pp. 1326-1329
Author(s):  
Zhao Kun Wang ◽  
Xiao Yang Zhang ◽  
Ming Yong Lai ◽  
Bao Ping Liu

In this<b> </b>paper, a model based on ELM is proposed to predict the air-conditioning load under drought conditions by analyzing the daily average air- conditioning load during the drought. The main meteorological factors that impact the air-conditioning load are considered in the model, and then the air-conditioning load under drought conditions can be predicted by training the samples by the single hidden layer feed forward neural network of ELM. Thus, the model is used to provide good theoretical basis for management on the demand side of power sector. Finally, an example is showed to prove that the curve of the air-conditioning load forecasting model and the curve of the actual cooling load of the power are almost consistent, and the prediction is accurate, reliable, and can be applied to other load forecasting.


2011 ◽  
Vol 19 (01) ◽  
pp. 57-68 ◽  
Author(s):  
MIGUEL PADILLA

Commercial multiple evaporators variable refrigerant flow (VRF) HVAC systems present many advantages such as being energy saving and the capability of adjusting refrigerant mass flow rate according to the change of high rises occurrence. This paper deals with an experimental control volume exergy analysis in a VRF air conditioning system. The experimental results show that the brunt of the total exergy destroyed in the whole system occurs in the outdoor unit, where the exergy destroyed in the condenser is more important. The values of coefficient of performance (COP) obtained for the tests increase as the system reaches operational conditions imposed in every indoor unit zone. The VRF system analyzed is highly sensitive to the action of the constant speed compressor. The use of an inverter compressor improves the system performance by adjusting the power consumption according to the cooling load in the evaporators.


2019 ◽  
pp. 9-14
Author(s):  
Євген Іванович Трушляков ◽  
Андрій Миколайович Радченко ◽  
Микола Іванович Радченко ◽  
Сергій Анатолійович Кантор ◽  
Веніамін Сергійович Ткаченко

One of the most attractive reserves of enhancing the energetic efficiency of air conditioning systems is to provide the operation of refrigeration compressors in nominal or close to nominal modes by choosing rational design cooling loads (cooling capacities) and their distribution according to a cooling load behaviour within the overall design (installed) cooling load band to match current changeable climatic conditions and provide close to maximum annual cooling capacity generation according to cooling duties. The direction of increasing the efficiency of outdoor air conditioning in combined central-local type systems by rationally distributing the heat load - cooling capacity of the central air conditioner into zones of variable heat load in accordance with current climatic conditions and its relatively stable value, i.e. cooling capacity required for further air cooling at the entrance to the indoor recirculation air conditioning system is justified. By comparing the values of the excessive production of cold and its deficit within every 3 days for a rational design heat load of the air conditioning system (cooling capacity of the installed refrigeration machine), which provides close to maximum annual production of cold, and the corresponding values of the excess and deficit of cooling capacity in accordance with current climatic conditions during July substantiated the feasibility of accumulating the excess of cooling capacity of a central air conditioner at low current loads and its use for covering cooling deficit at elevated heat loads through pre-cooling the outdoor air. It is developed a scheme of a combined central-local air conditioning system, which includes the subsystems for the outdoor air conditioning in a central air conditioner and the local indoor recirculated air conditioning.


2014 ◽  
Author(s):  
J. Fajardo ◽  
B. Sarria ◽  
M. Alvarez Guerra

This work has as object of study the energy of a river ship air conditioning system performance, using fiberglass, polyurethane or rockwool as insulation. Thermoeconomics Indicators based on second law of thermodynamics which take into account the quality of the energy and the cost of the exergy were used for research. It was observed that: (i) by increasing the thickness of the insulation the irreversibilities decreased, (ii) increases in the destroyed exergy increased generation of cooling load costs and (iii) costs per unit of exergy of heat load and area for the generation of cooling load and for investment in exergetic insulation, were minors for polyurethane.


2016 ◽  
Vol 819 ◽  
pp. 189-201
Author(s):  
M.F. Sukri ◽  
Md Nor Musa ◽  
M.Y. Senawi ◽  
H. Nasution

This paper presents a steady-state modeling and parametric study on the characteristics of cooling loads for an automotive air-conditioning system. The model enables the hourly prediction of vehicular cabin cooling load profile at various operating conditions and types of vehicle. The model utilizes correlations and equations proposed by ASHRAE and previous researchers available in the open literature. A validation exercise indicates that the simulation results are within 5% of published results. In the parametric study, the effects of five parameters: vehicle surface color, number of passenger, cabin temperature, vehicle speed and angle of front wind screen on the hourly cooling load profile on a daily basis are investigated. It was found that an increase in number of passenger and vehicle speed, darker vehicle surface color and lower cabin temperature will increase the cooling load. Meanwhile, the angle of front wind screen does not significantly influence the cooling load profile. In conclusion, the cooling loads profile can be used for air-conditioning system improvement/optimization exercise. The generic model can be used for complete vehicle air-conditioning simulation program towards energy-efficient air-conditioning system for better overall performance, especially in the early stage of vehicle development.


Sign in / Sign up

Export Citation Format

Share Document