Load Prediction Model Research Articles

Explainable XGBoost machine learning model for prediction of ultimate load and free end slip of GFRP rod glued-in timber joints through a pull-out test under various harsh environmental conditions

Explainable XGBoost machine learning model for prediction of ultimate load and free end slip of GFRP rod glued-in timber joints through a pull-out test under various harsh environmental conditions

Industrial steam consumption analysis and prediction based on multi-source sensing data for sustainable energy development

Centralized heating is an energy-saving and environmentally friendly way that is strongly promoted by the state. It can improve energy utilization and reduce carbon emissions. However, Centralized heating depends on accurate heat demand forecasting. On the one hand, it is impossible to save energy if over producing, while on the other hand, it is impossible to meet the heat demand of enterprises if there is not enough capacity. Therefore, it is necessary to forecast the future trend of heat consumption, so as to provide a reliable basis for enterprises to reasonably deploy fuel stocks and boiler power. At the same time, it is also necessary to analyze and monitor the steam consumption of enterprises for abnormalities in order to monitor pipeline leakage and enterprise gas theft. Due to the nonlinear characteristics of heat load, it is difficult for traditional forecasting methods to capture data trend. Therefore, it is necessary to study the characteristics of heat loads and explore suitable heat load prediction models. In this paper, industrial steam consumption of a paper manufacturer is used as an example, and steam consumption data are periodically analyzed to study its time series characteristics; then steam consumption prediction models are established based on ARIMA model and LSTM neural network, respectively. The prediction work was carried out in minutes and hours, respectively. The experimental results show that the LSTM neural network has greater advantages in this steam consumption load prediction and can meet the needs of heat load prediction.

Charging load forecasting of electric vehicles based on sparrow search algorithm‐improved random forest regression model

AbstractIn order to solve the problem that the current charging load forecasting accuracy is not high, it is difficult to simulate the actual charging load distribution of Electric Vehicles (EVs), and it is impossible to reasonably predict the future load, a charging load forecasting model based on Sparrow Search Algorithm (SSA) improved Random Forest Regression (RFR) is proposed. The SSA is used to enhance the ability of global optimization and local exploration. Combined with the advantages of the RFR model, such as low generalization error, fast convergence speed, and few adjustment parameters, the SSA was used to optimize the parameters of the decision tree number and the number of split nodes in the RFR, and the optimal value of the parameters is obtained, so as to obtain the optimal performance of the RFR. Firstly, based on the concept of travel chain and conditional probability distribution, the user's travel habits are described. Monte Carlo simulation method was used to simulate the driving, parking, and charging behaviours of a large number of EVs in different regions, so as to obtain the charging load of EVs in different regions. Then, a charging load forecasting model based on SSA improved RFR is established. Monte Carlo simulation results are used as sample data to predict the charging load of EVs in different regions. Finally, taking a certain area as an example, the experimental results show that the charging load prediction model based on Sparrow Search Algorithm improved Random Forest Regression (SSA‐RFR) can accurately predict the charging load of EVs in different regions, and the charging load of different regional types is obviously different. Compared with the RFR model and other literature models, the SSA‐RFR model has better prediction accuracy, which verifies the feasibility and superiority of SSA‐RFR model in EVs charging load prediction.

Comparative Study of Suspended Sediment Load Prediction Models Based on Artificial Intelligence Methods

Quantification of suspended load sediment is crucial for maintaining the ecosystem and quality of water/river bodies that serve as the habitat for many living organisms. Because the influencing factors are nonlinearly related to the suspended load sediment, it is a challenge to apply linear statistical models to predict accurately. To address such a problem, this study appliedartificial intelligence methods to simulate and predict suspended load sediment. The artificial intelligence methods are robust andcan handle adequately issues related to nonlinearity in modeling. In the present study, four artificial intelligence methods were developed to predict suspended sediment load distribution. The methods include a backpropagation neural network, group method of data handling, least squares support vector machine, and generalised regression neural network. In developing the respective models, drainage areas, river slopes, and length of rivers served as predictor variables while suspended sediment load was the response variable. The models were evaluated using the metrics of root mean square error, percentage root mean square error, uncertainty at 95%, root mean square error observations standard deviation ratio, and Legates and McCabe index. According to the results, the generalised regression neural network model achieved higher prediction accuracy than the other competing methods. The performance of the generalised regression neural network model can be attributed to its ability to calibrate and generalise appropriately to the training and testing data set. Hence, in practice, the generalised regression neural network model is proposed for suspended sediment load prediction for the study area which can be useful to policymakers and managers of water resources.

A control method combining load prediction and operation optimization for phase change thermal energy storage system

Due to the high latent heat and load-shifting capacity, phase-change thermal energy storage technology is an effective way to reduce energy costs under time-of-use electricity pricing. A favorable operation strategy is essential to exploit the advantage of the phase change thermal energy storage system. Previous studies on the operation strategy lack consideration of load prediction, which could reduce the matching degree of heat supply and demand. This study proposed a control method combing load prediction and operation optimization based on an electric boiler-phase change thermal energy storage heating system. A deep learning-based heating load prediction model was built; on this basis, an operation optimization method using dynamic programming was formulated subsequently. The proposed method was applied to a case building located in Beijing, China. By applying the proposed control method, the energy costs of the system were saved by 10% compared to the original operation strategy in the whole heating season, and the matching degree of heat supply and demand could be close to 100%, demonstrating that the method can provide users with cheaper and the best thermally comfortable heating. The proposed method is capable of achieving optimal operation of a phase change thermal energy storage system in practical applications.

Physics-guided LSTM model for heat load prediction of buildings

The thermal process of heating buildings is characterized by the multivariable, nonlinear, and large time lag. Accurately predicting building heat loads is crucial for regulating building heating systems. However, the current black-box method of heat load prediction has challenges such as extensive data usage and weak integration with the building's heat transfer process. To address these challenges, a physics-guided long short-term memory (LSTM) model for dynamic heat load prediction in the next seven days based on data-driven and mechanism analysis was proposed. The data for the model were collected from in-situ measurement and numerical simulation. Specifically, the thermal mechanism was introduced to the training loss function of LSTM and the selection of input variables. The influence of the adaptive data time step on prediction accuracy was investigated. The main conclusions obtained from the current data in this paper are drawn as follows. 1) Refining solar radiation intensity calculation for each orientation reduced mean absolute percentage error (MAPE) of heat load prediction results by 0.02%–1.8% compared to horizontal solar radiation intensity. 2) Improving the training loss function decreased MAPE of heat load prediction results by 0.04%–0.43% compared with the original loss function. 3) Adaptive data time step lowered the MAPE of heat load prediction results by 0.08%–1.8% compared with fixed time step. 4) Mean absolute errors corresponding to those three aspects decreased by 25 kJ–558 kJ, 6 kJ–157 kJ, and 41 kJ–1881 kJ. In the future, wider promotion and application of the prediction model will lead to better performance and be more beneficial to energy saving and carbon reduction of the heating system.

Global sensitivity analysis of fan coil air conditioning demand response—A case study of medium-sized office buildings

Air-conditioning demand response has drawn more attention as a means of reducing the ever-increasing peak load of the utility grid. However, the parametric analysis of the demand response performance is still insufficient, given a large number of design parameters of buildings, air-conditioning systems, climate zones, and demand response scenarios. In this study, four different sensitivity analysis methods, including the standardized regression coefficient method, the partial correlation coefficient method, the Morris method, and the extended Fourier amplitude sensitivity testing method, were adopted to screen the most important design parameters in determining the peak load, demand shedding intensity, and demand response potential of a medium-sized office building equipped with a fan-coil air conditioning system under the demand response control strategies of different global temperature adjustments. The results show that the four sensitivity analysis methods presented similar results of parameter screening, even with different degrees of accuracy and computational costs. For example, all the sensitivity analysis methods identified the occupied area per capita, fresh outdoor air per person, internal thermal mass, the safety factor of chiller capacity, and coefficient performance of chillers as the critical parameters for the demand shedding intensity. In addition, the critical parameters have similar rankings for their importance under different adjustment parameters, such as setpoint changes and duration time, in a demand response event. For critical parameters, their degrees of influence on demand shedding intensity are affected by climate zones, and the variation of coefficient for their sensitivity analysis indicators could be 20% higher than the least influential parameters. The study would benefit the design of air conditioning systems with load flexibility and prediction model simplification, etc.

Photovoltaic power generation and charging load prediction research of integrated photovoltaic storage and charging station

Aiming at the obvious randomness and intermittent problems of photovoltaic power generation output and charging load of photovoltaic storage and charging station, a photovoltaic power generation prediction model based on long short-term memory neural network and a charging load prediction model based on BP neural network are proposed respectively. In the photovoltaic power generation prediction model, the Pearson correlation coefficient is used to analyze the influencing factors of photovoltaic power generation power, and the factors with high correlation are used as the input variables of the long short-term memory neural network model. Then the data is divided into four types of seasons by K-means cluster analysis, and the processed data sets are put into the prediction model for training. In the charging load prediction model, the processed historical load data is put into the time series-based BP neural network for iterative training, meanwhile the weights and thresholds of the model are adaptively optimized. Finally, the simulation results verify the extremely high prediction accuracy of the proposed prediction model. The results also show that compared with PSO-BP, GA-BP neural network, this charging load prediction model has higher precision and stability.

Deep learning predictions of unsteady aerodynamic loads on an airfoil model pitched over the entire operating range

For the design and certification of wind turbines, it is essential to provide fast and accurate unsteady aerodynamic load prediction models for the whole operational range of angle of attack, up to 180° for vertical-axis and 90° for horizontal-axis wind turbines. This work describes a computationally efficient unsteady forces prediction model based on a deep learning approach, namely the bidirectional long short-term memory (BiLSTM) algorithm, for an airfoil pitched over the full operational range of angles of attack up to 180°. No model has been developed to capture the unsteady forces at high angles of attack. Novel features based on operating conditions and the steady polars of the airfoil are used as inputs for the BiLSTM model. Direct measurements of steady and unsteady forces on a NACA 0021 airfoil model were conducted at reduced frequencies up to 0.075 and a Reynolds number of 120 000 in an open-jet wind tunnel for model learning and testing. The unsteady forces vary significantly from the steady values at high pitching amplitudes and post-stall angles, which, if not accounted for when simulating wind turbine performance, would result in inaccurate predictions. Furthermore, measurements revealed the effect of unsteady vorticity development and shedding on aerodynamic forces under forward and reverse flow conditions. The BiLSTM model is capable of capturing the underlying physics of unsteady aerodynamic forces under extreme operating conditions.

A comprehensive scheduling model for electric vehicles in office buildings considering the uncertainty of charging load

The access of electric vehicles (EVs) has a profound impact on the operation and scheduling of building energy system. However, due to the fixed schedules of EV charging load, few studies concerned on the coordination strategy between EVs and other potential flexible load in office buildings. In this study, uncertainty sources of EV charging load are extracted from sensitivity analysis based on regression method, and then the uncertainties are introduced into the charging load prediction model based on Monte Carlo method. A optimal scheduling model embedded with the uncertainty of charging load, with the joint operation of building air conditioning system, photovoltaic system and EVs, is proposed for building demand response period. The results show that the operation strategies obtained from the proposed model embrace the advantages of small load variance, high flexibility, low operating cost and large photovoltaic power consumption capacity. The arrival non-guarantee rate is reduced from 12.23% to 6.55% and the charging non-guarantee rate is reduced from 24.03% to 9.39%. On the basis of EV orderly charging and the air conditioning pre-cooling, 7.57–61.67% of the 8:00 startup load can be transferred to night, and the original load variance can be reduced by 3.06–17.81%.© 2017 Elsevier Inc. All rights reserved.

Modeling and validation of a novel load model considering uncertain thermal disturbance in the district heating system

To accomplish precise heating, district heating systems need to understand the demand of heat users comprehensively. Most current heating load prediction models were built on energy center control, with only a few involving user-side control. Thermal disturbance factors differ depending on the type of heat user. This work established a novel heating load model (BHLP-UN model) for the user side in an existing district heating system. The uncertain thermal disturbances were identified and integrated into the base model (BHLP model) with hybrid mechanisms and deep learning methods. The model was calibrated and validated using actual operational data from different types of heat users during multiple heating seasons. The findings showed that the mean absolute percentage error of the BHLP-UN model was decreased by 53.512% −65.338% for different types of heat users compared with the BHLP model, which only considered the meteorological factors and the thermal inertia. Indoor temperature was set as an input variable to effectively analyze the time-varying property of user demand for temperature. When the indoor temperature was 22℃ / 18℃ (working /non-working), the annual load could decrease to 26.501% − 28.572%. The model achieves long-term load trends with short-term forecast accuracy.

An improved hybrid model for short term power load prediction

An improved hybrid model for short term power load prediction

Electric Vehicle Charging Facility Configuration Method for Office Buildings

With the advent of advanced battery technology, EVs are gradually gaining momentum. An appropriate decision-making method for the number of charging piles is in need to meet charging needs, and concurrently, to avoid the waste of infrastructure investment. In this study, an optimal charging pile configuration method for office building parking lots is proposed. With the determination of the design period of charging facilities, a charging load prediction model is established under a collection of charging scenarios. Taking the average utilization rate of charging facilities and the average satisfaction rate of charging demand as the objective functions, the distribution of the optimal number of piles is obtained with the genetic algorithm. The benefits of the configuration method are also explored under the building demand response process. The results show that the optimal configuration of charging piles in office buildings with different volumes have similar characteristics. When the design period is 5 years and 10 years, the comprehensive indicator of the utilization rate of the charging facilities and the satisfaction rate of the charging demand can, respectively, be improved by 8.18% and 17.45%. Moreover, the reasonable scheduling strategy can realize the load regulation response with a maximum load transfer rate of 25.55%.

Prediction of heating and cooling loads based on light gradient boosting machine algorithms

Machine learning models have been widely used to study the prediction of heating and cooling loads in residential buildings. However, most of these methods use the default hyperparameters, resulting in inaccurate prediction accuracy. In this work, based on hyperparametric optimization algorithms of random search (Random), grid search (Grid), covariance matrix adaptive evolution strategy (CMA-ES), and tree-structured parzen estimator (TPE), were combined with the light gradient boosting machine (LightGBM) model, to construct four hybrid models (Random-LightGBM, Grid-LightGBM, CMA-ES-LightGBM and TPE-LightGBM) for improved prediction accuracy of heating and cooling loads. The LightGBM model was trained using a dataset consisting of building features, cooling set points, and occupant behavior parameters. Feature selection was performed by a random forest-based feature selection method, which determines the input features of the load prediction model. The TPE- LightGBM model achieved the best prediction accuracy among all proposed models with a root mean square error (RMSE) of 0.2714, mean absolute error (MAE) of 0.1416, coefficient of determination (R2) of 0.9981, and mean absolute percentage error (MAPE) of 0.4699% for heating load prediction, and RMSE of 0.1901, MAE of 0.1394, R2 of 0.9924, and MAPE of 2.3509% for cooling load prediction. The proposed TPE-LightGBM model provides an efficient strategy for predicting heating and cooling loads, which can provide better energy efficiency measures at the early design stages of residential buildings.

A Short-Term Load Forecasting Model Based on Crisscross Grey Wolf Optimizer and Dual-Stage Attention Mechanism

Accurate short-term load forecasting is of great significance to the safe and stable operation of power systems and the development of the power market. Most existing studies apply deep learning models to make predictions considering only one feature or temporal relationship in load time series. Therefore, to obtain an accurate and reliable prediction result, a hybrid prediction model combining a dual-stage attention mechanism (DA), crisscross grey wolf optimizer (CS-GWO) and bidirectional gated recurrent unit (BiGRU) is proposed in this paper. DA is introduced on the input side of the model to improve the sensitivity of the model to key features and information at key time points simultaneously. CS-GWO is formed by combining the horizontal and vertical crossover operators, to enhance the global search ability and the diversity of the population of GWO. Meanwhile, BiGRU is optimized by CS-GWO to accelerate the convergence of the model. Finally, a collected load dataset, four evaluation metrics and parametric and non-parametric testing manners are used to evaluate the proposed CS-GWO-DA-BiGRU short-term load prediction model. The experimental results show that the RMSE, MAE and SMAPE are reduced respectively by 3.86%, 1.37% and 0.30% of those of the second-best performing CSO-DA-BiGRU model, which demonstrates that the proposed model can better fit the load data and achieve better prediction results.

A hybrid prediction method for short‐term load based on temporal convolutional networks and attentional mechanisms

AbstractAccurate power load prediction is an important guide for power system planning and operation. High‐ or low‐load prediction results will affect the operation of the power system. In recent years, deep learning technology represented by convolution neural network (CNN) and transformer has been proved to be suitable for power load prediction. This paper proposes a new short‐term power load hybrid forecasting model, called channel enhanced attention (CEA) and temporal convolutional network (TCN)‐based transformer comprehensive forecasting model. This method combines the short‐term feature extraction ability of TCN with the long‐term dependent capture ability of transformer for short‐term load forecasting. And the CEA designed in this study is added to improve the prediction accuracy. On the same dataset, the designed model predicts power load mean square errors of 0.056 and 0.146 for the next 24 h and the next week, respectively, which is 0.002 to 0.073 and 0.012 to 0.024 lower than the baseline model. The experimental results show that the hybrid short‐term power load prediction model proposed in this paper is significantly better than the existing methods. The predicted curve is in agreement with the actual charge change, which provides a good guidance for short‐term power load prediction.

A novel combined model for heat load prediction in district heating systems

Accurate heat load prediction is essential for improving the operational efficiency of district heating systems (DHSs). Numerous heat load prediction models have been proposed to improve the accuracy and stability. However, previous prediction models are limited by training data and hyperparameters and cannot obtain accurate and stable prediction results. This study proposed a combined prediction model to overcome the limitations of a single model. In the proposed model, four classical single models were selected as individual models for prediction. Subsequently, the minimum sum of squares of the combined prediction errors algorithm was used to calculate the weighting coefficient of each individual model. The final prediction results were obtained by combining the prediction results of the four models. The 1-h historical operation data from three heat substations in a DHS in northeast China were adopted to evaluate the performance of the proposed model. The multi-step prediction experiment results of the three datasets revealed that the average mean absolute percentage error values of the proposed model for the one-step, two-step, and three-step predictions reached 2.6%, 4.9%, and 7.0%, respectively; these percentages were significantly lower than those of the four individual models and three representative combined models. Therefore, the proposed model has better prediction performance than previous models in terms of accuracy and stability and can provide a theoretical basis for the precise regulation of DHSs.

Short-term load forecasting based on AM-CIF-LSTM method adopting transfer learning

Aiming at the unreliability of historical data for short-term load forecasting caused by the sudden change of power grid load under emergencies, a short-term load prediction method adopting transfer learning is studied. The proposed transfer learning method combines the attention mechanism (AM) with the long short-term memory network coupled with input and forgetting gates (CIF-LSTM) to construct the AM-CIF-LSTM short-term load prediction model. First, the variational modal decomposition (VMD) method is used to extract the trend component and certain periodic high-frequency components of the load datasets of the scene to be predicted and similar scenes. Subsequently, the AM-encoder/decoder learning model is established based on the trend component, and the AM learnable parameters are trained and transferred to the AM-CIF-LSTM model. Furthermore, inspired by the idea of classified forecasting, the load trend component and periodic high-frequency components under the required prediction scene are predicted by AM-CIF-LSTM and deep recursive neural network (DRNN), respectively. Finally, the load forecasting results are superimposed to obtain the load forecasting value. The experimental results demonstrate that the proposed method outperformed the existing methods in multiple accuracy indicators and could predict the rapid change trend of load in the case of insufficient data accurately and stably.

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.

Stepwise Optimization Method of Group Control Strategy Applied to Chiller Room in Cooling Season

Central air-conditioning systems account for the largest share of energy consumption in public buildings, wherein the chiller room is the main source. The current empirical strategies of chiller room group control have difficulty realizing integrity, timeliness, and equipment adjustment accuracy and lead to energy wastage. Therefore, the operation strategy optimization model is of great value for achieving energy savings. This study proposed a stepwise optimization method for central air-conditioning systems, which is divided into the chilled side (the chiller and chilled water pump) and the cooling side (the chiller, cooling water pump, and cooling tower). The optimal points of the two sides are calculated sequentially and integrated as the global optimization result. To construct the stepwise optimization model, mathematical power models of the equipment in use were established to provide mathematical support for objective functions, and a short-term load prediction model with a matching optimization step size was established for energy constraints. The TRNSYS simulation model was developed to verify the energy-saving effects of the stepwise optimization model according to the energy-saving rates of 6.41% and 13.56% attained in both cases. The stepwise optimization strategy can more effectively guide practical applications and provide another way of thinking with respect to the group control optimization of chiller rooms in public buildings.