Cooling Load Prediction Model Research Articles

Transformer based day-ahead cooling load forecasting of hub airport air-conditioning systems with thermal energy storage

The air conditioning system constitutes more than half of the total energy demand in hub airport buildings. To enhance the energy efficiency and to enable intelligent energy management, it is vital to build an accurate cooling load prediction model. However, the current models face challenges in dealing with dispersed load patterns and lack interpretability when black box approaches are adopted. To tackle these challenges, we propose a novel k-means-Temporal Fusion Transformer (TFT) based hybrid load prediction model. Specifically, the daily load patterns are grouped using an improved k-means clustering method that considers both input feature weights and dynamic time warping (DTW) distances. Additionally, the statistical features of the clustering output are inputted into the TFT. By further incorporating context information, the integration of data between different schema categories is achieved, thus reducing errors that may occur during the transition process. As a result, the prediction performance and interpretability are significantly improved. The Chongqing Jiangbei Airport T3A terminal is used as a case study, and experiments are conducted using cooling data from the No.1 energy station, as well as the airport traffic data and the meteorological station data. Results are compared with other mainstream models, confirming that the proposed day-ahead load forecasting model achieves improvements in several performance indicators, including MAE, MAPE, CV-RMSE, and R2, which are 384 kW, 3%, 5%, and 0.058 respectively.

Research on systematic analysis and optimization method for chillers based on model predictive control: A case study

Chillers are key components of heating, ventilation, and air-conditioning (HVAC) systems and consume large amounts of energy in commercial buildings. This study conducted field tests on the operation performance of chillers in large-scale commercial buildings and analyzed the typical problems influencing their energy efficiency. Subsequently, based on the field test results, a systematic optimization method for chillers was proposed based on external factors, internal factors, and synergy control between them. For external factors, the cooling load prediction model comprising a linear model based on the time series was constructed using historical operation data; the proposed model could be applied to guide the control of chillers to match the building cooling demand accurately with high energy efficiency. For internal factors, the energy performance model of the chiller was constructed using historical operation data, and the operation characteristics of constant-speed and variable-speed centrifugal chillers were compared and summarized. Then, using the cooling load prediction model and energy performance model of the chillers, a new control strategy was proposed to improve the energy performance under different operating conditions throughout the year. Compared with previous operation modes in 2018 and 2019, the energy saving rates reached 19.21% and 5.22%, respectively, achieving considerable energy-saving and CO2 emission reduction effects.

Research on Central Air Conditioning Systems and an Intelligent Prediction Model of Building Energy Load

The central air conditioning system provides city dwellers with an efficient and comfortable environment. Meanwhile, coinciding with their use, the building electricity load is increased, as central air conditioners consume a lot of electricity. It has become necessary to control central air conditioners for storage and to analyze the energy saving optimization of central air conditioner operation. This study investigates the energy consumption background of central air conditioning systems, and proposes an intelligent load prediction method. With a back propagation (BP) neural network, we use the data collected in the actual project to build the cooling load prediction model for central air conditioning. The network model is also trained using the Levenberg–Marquardt (LM) algorithm, and the established model is trained, tested, and predicted by importing a portion of the sample data, which is filtered by preprocessing. The experimental results show that most of the data errors for training, testing, and prediction are within 10%, indicating that the accuracy achievable by the model can meet the practical requirements, and can be used in real engineering projects.

Improved Harris Hawks Optimization with Hybrid Deep Learning Based Heating and Cooling Load Prediction on residential buildings

In digital era, energy efficient building remains a hot research topic because of increasing concern regarding their environmental impact and energy consumption. Designing a suitable energy efficient building based on their layout namely overall areas, distribution of the glazing areas, orientation, height, and relative compactness. Such components directly impact the heating load (HL) and cooling load (CL) of residential buildings. A precise predicting of load facilitates effective management of energy consumption and improves the quality of life. Lately, several studies have been implemented to predict the CL and HL. The most significant and challenging parts of predictive are defining the effective input parameter and developing a higher accuracy predictive model. The accuracy of predictive model based on machine learning algorithm must be enhanced by hybrid model. With this motivation, this article introduces an Improved Harris Hawks Optimization with Hybrid Deep Learning Based Heating and Cooling Load Prediction (IHHOHDL-HCLP) model on Residential Buildings. The major aim of the IHHOHDL-HCLP model is to determine the CL and HL to accomplish effective energy utilization. To accomplish this, the IHHOHDL-HCLP primarily pre-processes the raw data in two levels namely min-max normalization and polynomial equation. In addition, the HDL model involves convolutional neural network (CNN) along with long short-term memory (LSTM) and bidirectional long short-term memory (BiLSTM) for HL and CL prediction process. Finally, the IHHO technique was applied for fine-tuning the hyperparameters related to the DL model. The IHHOHDL-HCLP model has demonstrated maximum prediction results with low RMSE values of 0.00874 and 0.00821, respectively, when applied to HL and CL, respectively. The experimental result analysis of the IHHOHDL-HCLP model demonstrates the better performance of the IHHOHDL-HCLP model over other DL models.

A novel approach of day-ahead cooling load prediction and optimal control for ice-based thermal energy storage (TES) system in commercial buildings

Thermal energy storage (TES) is an effective method for load shifting and demand response in buildings. Optimal TES control and management are essential to improve the performance of the cooling system. Most existing TES systems operate on a fixed schedule, which cannot take full advantage of its load shifting capability, and requires extensive investigation and optimization. This study proposed a novel integrated load prediction and optimized control approach for ice-based TES in commercial buildings. A cooling load prediction model was developed and a mid-day modification mechanism was introduced into the prediction model to improve the accuracy. Based on the predictions, a rule-based control strategy was proposed according to the time-of-use tariff; the mid-day control adjustment mechanism was introduced in accordance with the mid-day prediction modifications. The proposed approach was applied in the ice-based TES system of a commercial complex in Beijing, and achieved a mean absolute error (MAE) of 389 kW and coefficient of variance of MAE of 12.5 %. The integrated prediction-based control strategy achieved an energy cost saving rate of 9.9 %. The proposed model was deployed in the realistic building automation system of the case building and significantly improved the efficiency and automation of the cooling system.

Prediction of functional zones cooling load for shopping mall using dual attention based LSTM: A case study

Prediction of cooling load for each functional zone is essential for implementing an efficient and economic energy management plan in shopping mall. While previous research has been concerned with the overall cooling load prediction of the shopping mall, which could not provide a reasonable control strategy for the on-demand cooling of functional zone in shopping malls. To address this, we present a novel approach to predict cooling load of functional zones in shopping mall which first uses gray relational analysis methods to screen the key influencing factors of cooling load considering the characteristics of cooling load for different functional zones in shopping mall. It then applies LSTM with dual attention mechanism to optimize the output of the prediction model. Feature attention extracts important features through the analysis of the relationship between historical information and input variables autonomously, while temporal attention improves the prediction accuracy by analyzing the importance of load of each historical moment to the prediction of time load. We evaluate the performance of the proposed approach using real data sets from a shopping mall, which locates at Xi 'an, in the north of China. Experimental results show that cooling load prediction based on function zones may not only better satisfy the cooling load demand of diverse zones, but also ensure that chilled water is available on demand for air conditioning systems. Meanwhile, the short-time zonal cooling load prediction model based on dual-attention-LSTM has increased prediction accuracy, generalization ability, and stability when compared to previous prediction models.

Short-term building cooling load prediction model based on DwdAdam-ILSTM algorithm: A case study of a commercial building

Traditional prediction methods of building cooling load fail to consider load changes caused by the time factor, which results in the failure of the best predicting effect and portability. To explore a more accurate cooling load predicting approach, in this paper, a new hybrid model based on decoupled weight decay Adam optimized improved LSTM neural network, denoted as DwdAdam-ILSTM, is proposed. First, the cross-correlation function is used to extract features with a high degree as the input data. Then, the decoupling weight decay adaptive moment estimation (DwdAdam) optimization algorithm is utilized to optimize the parameters of the improved long short term memory (ILSTM) prediction model. Next, the ILSTM is employed to establish a prediction model. Finally, to verify the validity and prediction accuracy of the proposed model, this paper selects the hourly cooling load data of a large commercial building in Xi'an as the research sample. Results of the experiment show that the Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), Mean Square Error (MSE), Variation Coefficient of Root Mean Square Error (CV-RMSE) and Range of Relative Errors (E-R) of DwdAdam-ILSTM predicting the cooling load are 2.2%, 2.74, 5.2%, 0.061, 0.216. Compared with three version variant LSTM neural networks, the MSE and E-R values of this model are decreased by 74%-53%, 42%-49%, and 69%-64%, which has better performance in accuracy and stability. It testifies that the proposed strategy is reliable and effective. Additionally, this model can provide important guidance for daily energy management and multi-energy scheduling of buildings, which is helpful to the application of cooling load prediction.

Building Cooling load prediction based on LightGBM

In building Heating Ventilation and Air Conditioning (HVAC) control systems, the HVAC must be controlled in real-time according to the actual building cooling load demand while minimizing energy consumption. Given that the building cooling load is affected by several factors and that historical data are sparse and have outliers due to sensor errors, existing prediction methods cannot accurately calculate the building's cooling load demand. Therefore, this paper first pre-processes the historical building cooling load data, fills in the missing data utilizing the KNNImputer algorithm, and then locates and replaces the outlier data through the 3-sigma method, solving the poor data availability problem. Moreover, to address the problem of small feature variety in historical data, this paper analyzes the time features and the climatic features separately. Specifically, our method splits the time features into several independent features for describing the impact of pedestrian flow on the building cooling load and expands the climatic features to increase the variety of the features that can be used for training and improving the model's prediction accuracy. Finally, the LightGBM algorithm is exploited to train the processed data and build a building cooling load prediction model. The proposed data pre-processing method and model training algorithm are validated through Python programming, with the results highlighting that the proposed method effectively improves the prediction accuracy of building cooling loads.

A cooling load prediction method using improved CEEMDAN and Markov Chains correction

Due to having the characters of non-linear and non-stationary, it is difficult to accurately predict the air conditioning cooling load. In order to predict the load more accurately, this paper introduces the method of Improved CEEMDAN algorithm and Markov chain correction. The Improved CEEMDAN algorithm decomposes the affecting parameters of cooling load into different components. With Pearson correlation analysis, the components with high correlation to cooling load are retained. The component prediction model of the cooling load is established, through Markov chain correction to weaken the effect on the accuracy by the model's uncertainty. The above method alleviates the influence of data fluctuation on the prediction accuracy of cooling load. Parallel computing is used in the cooling load prediction model, which enhances the model operation speed. The results show that the prediction accuracy of the initial data of the model is improved and the peak error of the prediction is significantly improved, compared with the traditional GA-BP network and PSO-BP network. The improved model prediction results are more in line with the application of practical problems.

Predictive model of cooling load for ice storage air-conditioning system by using GBDT

Cooling load prediction can provide reliable data support for the development of low energy consumption building. It is difficult for traditional models to achieve stable prediction accuracy under different load mode when the amount of data is limited. To resolve this problem, gradient boosting decision tree (GBDT) model based on selection of input variables was proposed to predict the cooling load of commercial buildings. First, the input variables that affect prediction cooling load were determined through correlation analysis on the original data set. Then, the Pearson analysis was done on the original data to get the optimal combination of input variables. Next, the ensemble learning technology was used to build GBDT model. In order to improve the prediction accuracy of the model, the grid search method was applied to determine the hyperparameter combination of GBDT. Experimental results showed that in the high-load mode, the mean absolute error (MAE) of the GBDT model was reduced by about 24% compared with the SVM, and was reduced by about 37% compared with the DNN model. Under the general load mode, MAE was reduced by about 30% and 53% respectively. Compared with the other models, the prediction accuracy of GBDT model has been significantly improved under different load mode. The proposed cooling load prediction model can provide reliable data support for the predictive control of buildings.

A generic prediction interval estimation method for quantifying the uncertainties in ultra-short-term building cooling load prediction

Ultra-short-term building cooling load prediction is useful for optimal operations of building energy systems. The uncertainties of predicted cooling loads affect the reliability of the optimal operations significantly. This paper proposes a generic prediction interval estimation method to describe the uncertainties quantitatively. The basic idea is that the residual of a predicted cooling load would follow the same distribution as the residuals of historical predicted cooling loads of the similar operating condition. A Chebyshev distance-based agglomerative hierarchical clustering approach is proposed to gather historical prediction residuals of similar operating conditions into the same cluster. A quantile-based approach is proposed to estimate the prediction interval of a predicted cooling load using the cluster of the most similar operating condition. To evaluate the proposed method, an artificial neural network-based cooling load prediction model is trained to predict the cooling loads of a real building. The prediction intervals of the predicted cooling loads are then estimated using the proposed method. The results show that the estimated prediction intervals of the predicted cooling loads are very reliable.

Ultra-Short-Term Building Cooling Load Prediction Model Based on Feature Set Construction and Ensemble Machine Learning

As the requirements for the optimal control of building systems increase, the accuracy and speed of load predictions should also increase. However, the accuracy of load predictions is related to not only the prediction algorithm, but also the feature set construction. Therefore, this study develops a short-term building cooling load prediction model based on feature set construction. The impacts of four different feature set construction methods—feature extraction, correlation analysis, K-means clustering, and discrete wavelet transform (DWT)—on the prediction accuracy are compared. To ensure that the effect of the feature set construction method is universal, three different prediction algorithms are used. The influences of the sample dimension and prediction time horizon on the prediction accuracy are also analysed. The prediction model is developed based on an ensemble learning algorithm utilising the cubist algorithm, and the performance of the prediction model is improved when DWT is used for constructing the feature set. Compared with other commonly used prediction models, the proposed model exhibits the best performance, with R-squared and CV-RMSE values of 99.8% and 1.5%, respectively.

Prediction of Cooling Load of An Energy Station based on GA-SVR

Taking an energy station as the research object, the external dry bulb temperature and load values at t-1, t-2, t-3 moments were selected as input parameters, and the load value at t moment was used as output parameters to establish the SVR(Support Vector Regress)cooling load prediction model, the key parameters of SVR are optimized by GA(Genetic Algorithm).The results show that the maximum absolute error between the predicted value and the actual value is 4.83 GJ/h, the maximum relative error is 9.2 %, the average absolute error is 1.25 GJ/h, and the average relative error is 2.4 %.

Development of a cooling load prediction model for air-conditioning system control of office buildings

Building cooling load prediction is of critical importance for achieving energy saving of air-conditioning system in office buildings. It not only benefits the energy-efficiency of the air-conditioning system, but is also important for the system stability. Many techniques have been developed for building cooling load prediction. These methods are normally arranged into three categories: regression analysis, energy simulation and artificial intelligence. Among them, the regression analysis methods are simple in mechanism and much practical for real application. However, traditional regression models are not sufficient to manage multi-parameter dynamic changes, and the outliers in prediction has not been well considered, making the accuracy of cooling load prediction not satisfactory. To promote the feasibility of regression methods for cooling load prediction of office buildings, an efficient regression model based on sensitivity analysis and the traditional autoregressive with exogenous (ARX) model (named as improved ARX model) is proposed in this paper. The improved ARX model keeps the constitution of ARX model, but uses specified variables that selected by sensitivity analysis. The quadratic terms of vital variables are included to reduce the impact of system non-linearity. A least square method is used to get the weight coefficient matrix for model training. Comparison studies are used to evaluate the prediction accuracy of the improved ARX model. The proposed model will largely improve prediction accuracy and more adaptive for real applications in the perspective of optimal control for HVAC systems.

Cooling Load Forecasting For Typical Office Buildings

The heating and cooling load prediction model of an office building is established by using DesignBuilder V5 software. Some boundary conditions are obtained through investigation, and the annual heating and cooling load of the building is calculated by simulation. The maximum value of the cooling load is compared with the actual operation, and the maximum relative error results are within 10%. The model can provide reference for the load prediction of other office buildings.

Development and Validation of a Cooling Load Prediction Model

In smart buildings, cooling load prediction is important and essential in the sense of energy efficiency especially in hot countries. Indeed, prediction is required in order to provide the occupant by his consumption and incite him to take right decisions that would potentially decrease his energy demand. In some existing models, prediction is based on a selected reference day. This selection depends on several conditions similarity. Such model needs deep analysis of big past data. Instead of a deep study to well select the reference day; this paper is focusing on a short sampling-rate for predicting the next state. So, this method requires less inputs and less stored data. Prediction results will be more close to the real state. In first phase, an hourly cooling load model is implemented. This model has as input current cooling load, current outside temperature and weather forecast to predict the next hour cooling consumption. To enhance model’s performance and reliability, the sampling period is decreasing to 30 minutes with respect to system dynamic. Lastly, prediction’s accuracy is improved by using previous errors between actual cooling load and prediction results. Simulations are realized in nodes located at a campus showing good adequacy with measurements.

Research on short-term and ultra-short-term cooling load prediction models for office buildings

Building cooling load predictions can be used to better understand energy demands and to improve the energy efficiency of HVAC systems. In this study, GA-SVR and GA-WD-SVR prediction models for short-term and ultra-short-term predictions for office buildings are established. The short-term cooling load prediction model is designed to outline an HVAC system’s operation strategies for the following day. The ultra-short-term cooling load prediction model is designed to inform building managers of the cooling load for the next hour and to adjust HVAC system operations in advance. An office building in Tianjin is used to train and evaluate the load prediction models. Meteorological data and one-day-ahead and one-hour-ahead cooling load records are used as model inputs. The prediction results indicate that the GA-SVR prediction model performs better for short-term cooling load prediction with MRE and R2 of 6.5% and 73.1%, respectively, while the GA-WD-SVR prediction model performs better for ultra-short-term cooling load prediction with MRE and R2 of 4.6% and 88.7%, respectively.

Prediction model of Cooling Load considering time-lag for preemptive action in buildings

As interest in automation technologies for maintaining occupants' comfort and reducing building energy consumption increases, Energy Management Systems (EMS) and Building Energy Management Services (BEMS) are gradually receiving more attention. However, the conventional methods associated with these systems cannot predict the future thermal load based on real time thermal load and climatic factors. This study shows that the external environment and building thermal load may not be the same due to the time-lag phenomenon. Consequently, prediction models that take the time-lag phenomenon into consideration are developed. In order to create the prediction models, meteorological data (mixed humidity) over 4 years (2011–2014) in Seoul were consolidated using the collected data by KMA (Korea Meteorological Administration) and mathematical equations used in energy analysis simulations. A cooling load prediction model per building size considering the time-lag phenomenon was proposed based on multiple regression analysis. It was found that there are some cases where energy could not be predicted as there is a different time-lag due to architectural characteristics and building thermal load conditions.

Estimating hourly cooling load in commercial buildings using a thermal network model and electricity submetering data

One major obstacle in Heating, Ventilation and Air Conditioning (HVAC) system Fault Detection and Diagnostics (FDD), retrofitting and energy performance evaluation is the lack of detailed hourly cooling load data. Cooling load measurement in commercial buildings is expensive and sometimes very difficult to implement. Detailed building simulation models, such as EnergyPlus, are too complicated to build and also must be calibrated. In this paper, an hourly cooling load prediction model, called the “RC-S” model, is proposed. This new cooling load calculation approach consists of a simplified thermal network model of the building envelope, a thermal network model for the building internal mass and the internal cooling load model from the submetering system. One existing RC model is introduced as reference model and three types of “RC-S” models are set up in this study. Genetic algorithm (GA) is selected to optimize the parameters in those models. Measurement data collected from a real commercial building and simulation data obtained from EnergyPlus model of the same commercial building are used to train and test the four models. The results prove that the proposed “RC-S” cooling load calculation method is more accurate than the existing RC model and much simpler than whole building simulation models. It can provide reasonable estimations of cooling loads for HVAC FDD and other performance evaluations.

An improved office building cooling load prediction model based on multivariable linear regression

The cooling load prediction of heating, ventilating and air-conditioning (HVAC) systems in office buildings is fundamental work for optimizing the operation of HVAC systems. In this paper, an improved multivariable linear regression model is proposed to predict the daily mean cooling load of office buildings in which three main measures, including the principal component analysis (PCA) of meteorological factors, cumulative effect of high temperature (CEHT) and dynamic two-step correction, are used to improve prediction accuracy. The site measured cooling load of two office buildings in Tianjin is used to validate the model and evaluate the prediction accuracy. Meanwhile, four contrast models with one or two of the three measures are also built. A comparison among the models proves that a combination of the three measures could effectively improve the prediction accuracy. The predicted load of the proposed model has acceptable agreement with actual load, where the mean absolute relative error is less than 8%.