Uncertainty Analysis of Weather Forecast Data for Cooling Load Forecasting Based on the Monte Carlo Method

Jing Zhao,Yaoqi Duan,Xiaojuan Liu

doi:10.3390/en11071900

Abstract

Recently, the cooling load forecasting for the short-term has received increasing attention in the field of heating, ventilation and air conditioning (HVAC), which is conducive to the HVAC system operation control. The load forecasting based on weather forecast data is an effective approach. The meteorological parameters are used as the key inputs of the prediction model, of which the accuracy has a great influence on the prediction loads. Obviously, there are errors between the weather forecast data and the actual weather data, but most of the existing studies ignored this issue. In order to deal with the uncertainty of weather forecast data scientifically, this paper proposes an effective approach based on the Monte Carlo Method (MCM) to process weather forecast data by using the 24-h-ahead Support Vector Machine (SVM) model for load prediction as an example. The data-preprocessing method based on MCM makes the forecasting results closer to the actual load than those without process, which reduces the Mean Absolute Percentage Error (MAPE) of load prediction from 11.54% to 10.92%. Furthermore, through sensitivity analysis, it was found that among the selected weather parameters, the factor that had the greatest impact on the prediction results was the 1-h-ahead temperature T(h–1) at the prediction moment.

Highlights

In recent years, heating, ventilation and air conditioning (HVAC) systems have become important elements in office buildings and are responsible for around 40% of the energy use in office buildings, which means a great energy-saving potential [1]
The data-preprocessing method based on Monte Carlo Method (MCM) makes the forecasting results closer to the actual load than those without process, which reduces the Mean Absolute Percentage Error (MAPE) of load prediction from 11.54% to 10.92%
Precise load forecasting is the basis of the optimization of HVAC system operation, which is conducive to formulate an operation strategy according to the load change and can lay the theoretical foundation for enhancing the thermal comfort and reducing the energy consumption of office buildings

Summary

Introduction

In recent years, heating, ventilation and air conditioning (HVAC) systems have become important elements in office buildings and are responsible for around 40% of the energy use in office buildings, which means a great energy-saving potential [1]. The operation management level of HVAC systems is generally low, and the refrigeration capacity of the equipment does not match with the actual demand, resulting in a large energy consumption. Meteorological parameters play a very important role in the dynamic cooling load, which has a great influence on the actual energy consumption of a building. In the relevant literature on building load forecasting, various prediction models are proposed for load forecasting and related research. Xia and Xiang et al [2] proposed a prediction model based on a radial basis function (RBF) neural network to forecast a daily load, which mainly took some weather parameters into consideration including temperature, humidity, wind speed, atmospheric pressure and so on. The forecasting results illustrated that the model has better performance compared with the Back

Methods

Results

Discussion

Conclusion