Air-Conditioning Load Forecasting for Prosumer Based on Meta Ensemble Learning

Abstract

Accurate and reliable prediction of airconditioning load plays a significant role in prosumer energy management system (EMS), because airconditioning load accounts for a large proportion of the building's total energy consumption. This paper proposes a new meta ensemble learning method to realize short-term prediction of airconditioning load for prosumers. This method is a hybrid of meta ensemble learning and stacked auto-encoder (SAE). First, we design multiple different forecasting structures based on SAE to achieve point prediction of airconditioning loads. SAE is used to learn the deep features in airconditioning load data. Second, a new meta ensemble learning prediction model is proposed. Meta ensemble learning is used to learn the nonlinear features and invariant structures in data, and determine the coefficients of each SAE-based point forecaster. Finally, the prediction results of each point forecaster are aggregated and integrated to estimate the final airconditioning load prediction result. Airconditioning load data from a commercial building in Singapore are used to validate the feasibility and effectiveness of the proposed method, demonstrating that the proposed meta ensemble learning method is attractive in prosumer energy management. INDEX TERMS Airconditioning load, load prediction, meta learning, ensemble learning, prosumer energy management system.

Application of Deep Learning Model in Building Energy Consumption Prediction

“…Chen et al found that the most commonly used physical model software is EnergyPlus, TRNSYS, DeST, and so on. [ 10 ]. Liu et al found that when using the concepts of physics to create the best working model in the house, the physical estimation model needs to permeate many negative buildings, the environment is not good, and the design time is long [ 11 ].…”

“…Formula (1) shows the function of activation function, and different activation functions will make the neural network have different characteristics. Several commonly used activation functions are described below: ladder function formula (5), sigmoid function formula (6), tanh hyperbolic function formula (8), and ReLU modified linear unit function formula (10). Ladder function is relatively simple and generally used for teaching.…”

Deep Learning with Dipper Throated Optimization Algorithm for Energy Consumption Forecasting in Smart Households

“…Using an LSTM and auto-encoder persistence model to account for uncertainties and make predictions for complicated meteorological variables [27], successfully forecasted photovoltaic electricity for the following day. To enhance prosumer energy management, an air conditioner's energy usage was estimated using a machine learning model with meta-ensemble and stacked auto-encoders [28,29]. Using a mixed ensemble deep learning model based on a deep belief network, the authors of [30] could predict low-voltage loads with high certainty and uncertainty.…”